During her crazy adventure to Finland for the Persuasive Technology conference (more on that here), CLM’s Amy Greer met a fascinating gentleman from the UK named Richard Sedley. Richard works at a UK-based agency called cScape managing the Customer Engagement group. Amy introduced me to Richard when she returned and, in an attempt to brush up on my transoceanic diplomacy skills, we’ve been conversing ever since.

Recently, Richard graciously asked CLM to participate in the Customer Engagement Survey he runs in conjunction with E-Consultancy. Having seen the people who participated in the past surveys (like Andy Beal and Avinash K), we jumped at the chance. You can see the summary report from the previous surveys here.
So I’m putting out the call to all companies (including agencies) to take the survey. It only takes a few minutes. I found it to be an eye-opening experience.

Take the survey 

SMX East is happening Oct 6-8 in NYC, where I’ll be speaking on SEO & Usability. This promises to be one of the biggest and best Search conferences.The friendly folks who run the SMX events are letting speakers offer a $150 discount off all-access passes. Just comment on this post or email info (at) closed-loop-marketing.com if interested and we’ll send you the discount code. Please look me up if you’ll be there. SMX East site

The Iceberg’s Tip

Since their inception, search engines have relied on the visible to locate relevant content. Visible text, to be precise. And not just any visible text, either - the text had to be accessible and readable to a web crawler. That is, it couldn’t be inside an animation, image, script, video, or a wide assortment of other file formats. It couldn’t be stored in a ‘deep web’ database such as the CDC or USGS, one that was reached only via an active user query. And it certainly couldn’t be spoken.

This meant that, for all the blow-your-mind number of visible web pages out there on the web, all this time only a tiny fraction of the available content has been indexable and searchable by search engines. In 2001 the company BrightPlanet estimated in their white paper “The Deep Web: Surfacing Hidden Value” that public search engines made only 0.03% of the total web content available to searchers. That’s tiny. And this estimate wasn’t even considering content hidden in images, audio files, and video. Like a giant iceberg of data and content, the majority of the web remained - and still remains - invisible to search. But this is all changing.

Making sense out of sound

Google’s beta release of Gaudi, its audio indexing tool heralds to the wider public a profound shift in the search environment. Why? Hasn’t audio search been around for years now? Actually, no. Not this way.Yahoo’s audio search tool and video search tool have been around for quite a while, yes, along with Google’s video search tool. But these tools merely help you locate audio and video files - they don’t index the meaningful content within those files. Words that are spoken inside these files couldn’t be indexed. Neither could faces, locations, or other visual content information. Which means in order to attach real meaning and searchability to the files they had to be tagged with identifying information or else have an accompanying summary. Text, in other words.

Instead of relying on tags, meta data, or transcripts, Gaudi is capable of indexing the actual spoken content within audio and video files. If a speaker says the word “orbit” in a video, a search for the word “orbit” using Gaudi will be able to locate that video file, even if it has no tags, meta data, or nearby text containing that word.Why is this important? Aren’t tags, meta data, and transcripts good enough?

A Question of Quality

It shouldn’t surprise anyone to learn that most images, videos, and audio files on the web have insufficient or inaccurate information associated with them.Here are some of the reasons why:

• It takes time, effort, money, and planning to add ALT properties to images
• Same goes for adding meta data to videos
• attempts to leverage “crowd intelligence” to tag these files often results in inaccuracies or less-specific information (i.e. “dog” instead of “poodle”)
• tagging and meta data can only summarize content
• Many sites don’t have room for - or don’t want to display - complete transcripts or even text summaries of video and audio files

So, to summarize, we humans are fairly messy, lazy, and careless when it comes to identifying things - at minimum, underfunded - and that’s not going to change. Much better to develop an automated process to make sense of it all for us, calculate relevance, and serve it up in a familiar search results environment.

But Can You See Me? I Mean, Really See Me?

Behind the scenes, Google, Yahoo, and a wide range of other organizations have been hard at work figuring out how to crack the visual, as well as audio, indexing challenge. Google foreshadowed their visual video indexing capability in a blog post on June 14, 2007, saying,

“The technology extracts key visual aspects of uploaded videos and compares that information against reference material provided by copyright holders.”

The main purpose of this effort was to identify copyright violations on the YouTube platform, and it followed on the heels of Google implementing audio fingerprinting technology on their AudioSwap platform. Once developed, however, the technology has far-reaching capabilities and implications for extending visual search capabilities.

One intriguing recent development on the visual search side of things - and there are many - is the creation of face-recognition software that works even with low-quality images and video clips, like those found all over YouTube and the web. Developed by researchers at Carnegie Mellon, this face-recognition system solves one of the challenges of extracting information from image and video files - that of file resolution, poor lighting, non-controlled subject aspect (which way the subject’s face is turned), and overall image quality. Combined with other visual recognition software and image annotation methods, this technology will likely lead to widespread automated indexing of at least a portion of that other 99.97% of the web.

OK, We’ve Got the Kitchen Sink - Now What Do We Do With It?

But information is one thing — making it meaningful is another. With access to more data from more sources than ever, the question is what good will it do us? How do we tie together related information that’s stored in a variety of formats, locations, and languages? How do we not only locate and index this data but correlate it in useful and intelligent ways? Well, now we’re talking about the semantic web, called by some - especially those like myself who’re really tired of talking about Web 2.0 - “Web 3.0″. And that, my friends, is the topic for a whole other blog post.

Let me share with you how one PPC advertiser achieved…

• An ROI increase from 77% to 323%
• A conversion rate increase from 6.15% to 8.39%
• A CTR increase from 6% to 8%

…solely through match type testing and refinement.But first some background…Match types have come under more and more scrutiny in the past 2 years, given the coming of Google’s Expanded Broad Match and Automatic Match types. In fact, the unmonitored wreckless use of Broad Match was the subject of a presentation I made this year at SMX West’s Avoiding PPC Pitfalls search session.The fact that I presented Broad Match as a ‘pitfall’ actually does not mean that I am adamantly against the use of Broad Match. It can be a powerful tool for exploiting the long-tail and increasing sales and leads.But Broad Match (and Yahoo’s Advanced Match for that matter) can all-too-easily become a major pitfall, if left unchecked.Like any other PPC lever at our disposal, match types do not lend themselves to any sweeping generalizations - ‘always broad match’, ‘never broad match’, ‘only use exact match’, etc. Instead, like any other element of PPC you should continually test, monitor and adjust with the end goal of improving ROI and in turn increasing the lead or sales volume for your company.Even though my SMX presentation was over 6 months ago, this issue of match types has resurfaced and spurred me on to write this blog post for 2 closely related reasons.First, we’ve followed our own test-monitor-adjust advice and discovered some more interesting match type learnings since SMX West that I’d like to share with our readers. I’ll get to that in a bit…The second reason I revisit this topic now is that PPC Hero (one of my favorite PPC blogs which I HIGHLY recommend subscribing to, by the way) posted 2 match type related articles in the past month. The first, A Poor PPC Account Structure Will Make Your Campaign Suffer, described some oddities that Hanapin Marketing discovered upon inheriting a PPC account. One of these oddities (which turns out not to be an oddity after all) was that the campaigns were broken into ad groups according to Broad, Phrase and Exact match types.

The second major offense, and this strategy puzzles us, is that they had match types separated into different ad groups. For example, one keyword would appear in three different ad groups; one for broad, one for exact and one for phrase. Using match types can help determine user intent and you can use this data to focus on the match type that works best for your audience. But separating the match types doesn’t make much sense.

This account structure is actually the same highly effective manner in which Closed Loop Marketing has conducted some of its match type testing! In fact, apparently other PPC Hero readers have tested similar tactics and alerted the author that separating different match-typed keywords into unique ad groups was in fact a viable match type strategy.

Admirably, Joe posted a 2nd article less that a week later: Two Match Type Strategies That Can Enhance Your PPC Performance:

Last week I wrote an article about an account that we inherited that had been constructed and managed poorly. One of the strategies I called into question was separating keyword match types into unique ad groups within Google AdWords. However, we had quite a few comments from our readers stating that this is a strategy that they employ frequently with great results.This is one of the best things about our blog: I’m always learning how different people manage their PPC campaigns! After conversing with our commentators I thought I would give everyone a summary of what has been discussed.

So, in light of all of this — our match type testing results and the recent PPC Hero’s articles that directly tied in — I would like to share with you how one of Closed Loop Marketing’s clients has increased its ROI more than 4X by refining its match types (finally, you’re thinking).One of our clients that we provide PPC auditing services for had historically been broad matched across-the-board for nearly all of their search terms. Upon auditing their campaigns in search for areas of improvement, we discovered literally hundreds of thousands of dollars of wasted spend that had resulted from irrelevant broad matched iterations. It was the worst case of broad-match-gone-bad that we had ever seen.In the midst of all of this poor broad matching that had occurred within their account, their coverage levels for their most strategic keywords averaged only 30-50%! So essentially, this client’s ads were dark over half of the day for the keywords that were their largest revenue drivers, because their budget was being eaten up by keywords that had nothing to do with their offerings!Needless to say, one of our many recommendations was to test other match types, tracking each for revenue and ultimately ROI. Couple this with building out a robust negative keyword list and we’d be on our way to seeing some major improvement.Curiously, this was met with a lot of resistance by those actually managing the campaigns in question. They felt quite adamantly that everything should remain Broad Matched. So being responsible advertisers, we all agreed to put it to the test.Much like what is described above by PPC Hero, a subset of keywords were chosen, all 3 match types were applied to each keyword and then those keywords were split into Broad, Phrase and Exact match ad groups. Aside from the match types, all else was equal - same ad copy, same landing pages, same bid strategies. During the first half of the test, with all bids set to target #1 positioning, the results were as follows:As you can clearly see, the ROI, conversion rates and CTR for the Exact and Phrase matched keywords are significantly higher than their Broad Matched counterparts. The first thought is to toss out Broad Match all together. But…During the 2nd half of the test, in order to see if Broad Matched keywords could be salvaged and still prove some value (i.e. be made to be profitable), the Broad Matched keywords were bid down to position 3. Exact and Phrase Matched keywords remained in position 1.The results:The Broad Matched keywords increased from a measly 14% in the #1 position, to 261% in the #3 position! In essence, even though the Broad matched iterations of the keywords weren’t as valuable as the Phrase and Exact matched terms, by placing lesser value on them (i.e. placing lower bids) they were capable of producing revenue at a profitable level.Now don’t get me wrong, this is not the end of the story. The next task is to apply this to a larger set of keywords and continue to refine. In fact, the client is now on Round 3 of this match type testing and has since found that Phrase Matched terms needed to be bid down to position 2 in order to maintain profitable ROI. For the time being, the strategy stands as Exact Match ad groups being bid to position 1, Phrase Match position 2, Broad Match position 3.This particular position-based, match type strategy is just one lesson learned for one of our clients. Many of our other clients aren’t using Broad Match whatsoever because it just simply hasn’t proven to be advantageous for them at their current budget levels.And this is why we need to avoid making any sweeping generalizations. All campaigns are individual - test, monitor and adjust.I highly encourage all PPC advertisers to conduct similar tests on their keywords, as match typing is an often overlooked strategy that if refined can be a dramatic source of increased revenue and ROI.

More than a year and a half ago, my CLM colleague Sandra Niehaus and I attended the Second Annual Persuasive Technology conference at Stanford University in Palo Alto, California, hosted by BJ Fogg and his team at Stanford’s Persuasive Technology Lab.

At that time, I had only a rudimentary idea about what Persuasive Technology was and what its applicability to our company and my work might be.

Fortunately, I have a boss (and good friend) who encourages us CLM-ers to explore avenues that may not initially seem to be related to online marketing because of the opportunities such varied experiences and perspectives can offer.

And so we went. A summary of our experience at Persuasive Technology 2007 can be found here.

Little did I know that attending last year’s Persuasive Technology conference would result in attending this year’s conference in June (yeah, I’m tardy in publishing this post) – but this time it was in Oulu, Finland.

Yes, Finland.

If I would have flown straight there from Sacramento, it would have taken 21 hours and 3 layovers along the way. Fortunately, one of the stops was Paris.

So I took a few days’ vacation.

But that’s another story.

Needless to say, if you ever get the chance to take a solo trip to Paris for 5 days in late May, do it.

But enough about my vacation philosophies. Let’s talk about the conference.

Once in Oulu, I was part of a group of 113 attendees representing 20 countries at Persuasive Technology 2008. From the conference program, ‘PT 08’ is described as “a gathering of people interested in how software and related technologies influence people’s attitudes and behaviors.”And it was, but more about that in a second.

First, a few quick things about Oulu:

With a sun that never fully sets in the summer, free wireless throughout the city, a city government that welcomes its conference guests by hosting a reception in City Hall, and a Great Barn of the Maikkulankartano Mansion where I ate reindeer for the first time (yep, reindeer), Oulu is an enchanting place.

Oh, and it’s also home to the Air Guitar World Championships.

Wish I could have stayed for that. Next year, maybe.

Of note: Oulu also boasts large signs warning of very scary-looking mosquitoes. Fortunately for me, my encounters with them were brief – and involved only a little running.

When not dodging mosquitoes, eating reindeer or watching the sun not set at 4:30am, I immersed myself in Persuasive Technology 2008.

Part of this immersion also included being on the official Organizing Committee with an international crew of talented students and researchers as well as sitting on the Ethics Panel with Robert Biddle, Magnus Bång and Peter Øhrstrøm. I was honored to participate.

Some of the other topics covered this year at Persuasive Technology 2008 included:

• Online persuasion in social network systems, such as Facebook
• Website credibility, specifically regarding ‘active trust’ and behavioral intent
• Self-persuasion and ‘just in time’ motivational messages
• Systems for improving text passwords through persuasion
• Looking at persuasive technology through the paradigm of ‘Designing with Intent’

The conference program includes a full list of all the presentations, and the official Persuasive Technology 2008 conference proceedings can be reviewed and purchased here.

While this year’s conference tended to focus more on the academic and theoretical aspects of Persuasive Technology, a few presentations set the stage for combining the rigor of academia with ‘real-world’ industry applications.

This potential synergy between academia and industry was what I was looking for in attending this year’s conference.

Two of these presentations are featured below.

Persuasion for Stronger Passwords
Presenter: Alain Forget
School of Computer Science
Carleton University, Ottawa, Canada

Paper Summary:
Amidst a fervent and heated discussion amongst some conference attendees, Alain Forget from Carleton University introduced a ‘password creation mechanism’ that uses Persuasive Technology to influence users to create stronger passwords.

The premise of Alain’s paper rests on the fact that “text passwords are the ubiquitous method of authentication used by most people for most online services” (excerpted from paper’s abstract) and yet many users still select weak passwords.

Why do they do this?

Some security professionals postulate that the problem of weak passwords can be attributed to a lack of user effort and motivation. Others suggest that users create weak passwords due to a misunderstanding of security threats and / or the limitations of human memory for highly secure passwords (read: loG6%47t).

Yeah, I wouldn’t remember that either.

Thus, to address this problem, Alain and his co-authors consider how principles of Persuasive Technology can help users create stronger passwords that are nonetheless memorable, through the use of their Persuasive Text Passwords Prototype system.

Alain’s paper, Persuasion for Stronger Passwords, is now accompanied by another, more recently published paper, Improving Text Passwords Through Persuasion, from his presentation at SOUPS (Symposium on Usable Privacy and Security) in July of this year.

Applicability to Industry:
From an online marketing perspective, a website visitor’s decision about whether or not to engage in an ecommerce purchase can be greatly influenced by the visitor’s impression of the security of that website.

And if that user is in the habit of selecting weak passwords, there is a greater chance of an information breach, thus potentially resulting in a decreased level of trust for that site and for ecommerce in general.

As a result, research like that which is presented by Alain and his colleagues in the papers featured above, provides a unique opportunity for an intersection between industry and academia in the promotion of selecting more secure passwords through persuasive mechanisms.

Nice work, Alain and team. Looking forward to seeing the continuation of your work.

Design with Intent: Persuasive Technology in a Wider Context
Presenter: Dan Lockton
School of Engineering and Design
Brunel University, Uxbridge, West London

Scheduled for the less-than-ideal afternoon slot on the third (and final) day of Persuasive Technology 2008, Dan Lockton’s Design with Intent presentation was nonetheless one of the rock star presentations of the conference.

Paper Summary:
Dan’s paper introduced the field of ‘Design with Intent’ by explaining that Persuasive Technology (PT) is “an example of design intended to result in certain user behavior,” and that PT is “strategic, with defined, behavioral outcomes in mind” (taken from paper’s Introduction). Thus, PT fits into the wider context of Design with Intent (DwI).

What then followed in Dan’s talk was a series of images, photos, and examples of DwI in the broader world with a succinct and energized contextualization.

(You really must watch Dan’s Lockton’s SlideShare presentation.)

From ‘defensive design’ to ‘poka-yoke’ (from the Japanese for ‘mistake-proofing’) to ‘eco-design’, Dan effectively and enthusiastically brought the wider world of industry, philosophy and architecture as viewed through the paradigm of Design with Intent, directly to the door of Persuasive Technology.

You should have heard the applause.

Though there is no way I can effectively represent his work here, trust me when I say the brilliance of Dan’s work also lies in its potential. Though seasoned, he is still on an upward trajectory in his career and his contribution to DwI – and his work’s potential for broader applications – means he is ‘one to watch.’

I also encourage you to check out his site at: http://danlockton.co.uk/

Applicability to Industry:
In order to frame the significance of Dan’s work in the context of industry applications, specifically the online industry, I have excerpted some of Dan’s work on his What is Design with Intent? page:

“From the layout of shopping malls to digital rights management, our everyday lives are full of examples of products, systems and environments which have been designed to influence, guide or control – shape – our behaviour, using a wide range of techniques: technological, physical or psychological.

Intended commercial benefit is a major reason behind this; intended social benefit can be too (though it’s often fraught with politics). Sometimes the behaviour-shaping is helpful to the user; sometimes it’s serving someone else against the user’s best interests. Sometimes it’s trying to get the user to do something; sometimes it’s trying to stop the user doing something.

The common factor to all of this is intent on the part of the designer/engineer/planner or his or her corporate/political masters. This is Design with Intent: strategic design intended to result in certain user behaviour.”

Dan continues by stating:

“…Design with Intent techniques can be abstracted to a set of possible ‘tools’ (both physical and psychological) which can then be applied to other situations where a certain target behaviour is desired on the part of the user.”

Thus, the possible intersections between the work that Dan and others like him are doing and the fields of web usability, ecommerce, and conversion-based marketing become clear – and ripe with potential.

What remains, then, is for those of us on the side of industry to seek out opportunities to pursue those intersections.

In the meantime, best of luck to you, Dan Lockton, and others in your field. You’re doing great work.

Persuasive Technology 2009

Next year’s Persuasive Technology 2009 will be held at Claremont Graduate University in Claremont, California. More details will be provided as they are available.

Until next time …

This post is a continuation of a topic I began a few weeks back, about the cumulative impact that small, often overlooked interactions can have on the relationship between a company and their audience.

Earlier I discussed Thank You pages (see Optimizing the Thank You Page). Today I’ll briefly discuss error messages, cover some optimization basics, and examine two examples.

It’s True - Error Messages Always Come at a Bad Time

There’s a short story behind this saying. Once I was reviewing a potential client’s site with a colleague, and we triggered an error message while trying to use the product search functionality.

“Wow,” I commented. “Good-bye to that sale - that is not a good time to have an error message come up.” My colleague, who’s a bit of a smartass and hence shall remain nameless, smiled and replied, “You know, error messages always come at a bad time.”

Which is my long-winded way of pointing out that when a visitor gets an error, they’re not having fun on your site anymore. Their path was clear, their task was achievable, and now with no warning at all they’ve been thwarted in their ambition. Irritation sets in as they attempt to understand what happened, what to do next, and - most importantly from a business’s point of view - whether to continue trying at all.

And here, at this touch point, is your chance to make or break the relationship with that visitor.

The Message Might Be the Error

Visitors aren’t shocked when errors happen, it’s common enough on the web. Most will shake their heads and muddle through if given a clear, friendly direction. The point is, a decent error message can save - or even improve upon - your company’s relationship with the visitor. A poor error message, on the other hand, can completely trash that relationship, destroying trust and credibility.

Many error messages don’t even live up to common courtesy standards. Anyone remember this classic, from the 1996 era of browser battles?

“Your browser sucks! Get Navigator!

Pretty funny to developers, sure. But obviously not very likely to close a sale.

How Error Messages Become a Problem in the First Place

So, if it’s so obvious how important this is, why do so many sites have such rotten error messaging? Aren’t the site owners paying attention? Surely they didn’t approve those… right?

Probably not. And the blame falls on many shoulders.

Shoulder #1: The Developer Mindset

Most developers shouldn’t write public error messages, and to be honest most don’t want to. They get stuck with the task because they’re the only ones who know how to get the messages to display at the right time.

During my early years learning how to program, it took only one low project grade to teach me the value of good debugging messages. The messages I used were cryptic and highly personalized, meant to help me track the progress of a program’s test run and locate the point of failure in the code: “gonogo loop entered”, “success_count value now = x “, “charTwist function called”. Beautiful, aren’t they? Very helpful to me and possibly other programmers, but not very meaningful to anyone else.

This, then, is the mindset most developers bring to public-facing error messages, and why they shouldn’t write them: that error messages are meant to help the developer, not the site visitor.

Shoulder #2: The Site Development Process

How often is error messaging called out as an approval item during site development planning? Not often enough, judging by the evidence out there.

Partly it’s a problem of roles. We already decided that developers shouldn’t write error messages, so who should? Marketers? Content writers? A usability consultant? This should be determined and specified in the development plan.

Another part of the problem is the hidden quality of error messages. They’re part of a site’s dark underbelly, usually tucked away from the client’s eyes during reviews. Who sees them? The Quality Assurance team (assuming there is one) who probably don’t care how the message reads as long as it displays when it should. Oh, and actual site visitors, once the site goes live. This scenario could be avoided by treating error messages as legitimate site content that requires as much scrutiny as the home page copy.

Enough said. Let’s get to the guidelines, already! As usual, I present them alongside the visitor question or reaction the guideline is meant to address.

Guidelines for Error Message Optimization

Visitor Question	OPTIMIZATION GUIDELINE
What is this new thing I’m seeing?	First of all, make sure the error message CAN be seen. It should display in an obvious location and stand out against the page.Identify the message as an error message.”Error”, or “Oops!” work well, depending on your level of formality.
What just happened?	Tell the visitor what occurred, or what is happening now. Be brief, descriptive, and clear. Don’t try to explain the inner details, just a summary.”…your account has been locked for your security.”
Did I do something wrong?	Tone is important here. There’s no upside to blaming the visitor and reducing their confidence level. Take all the blame for the error, and apologize. Yes, maybe the visitor did something really stupid, but your site and pride should be able to handle it.”We’re sorry! Our calculator seems to be having a problem.”NOT: “You idiot! Don’t you know you’re not supposed to enter letters into a calculator?”
This is really annoying. You better be nice.	Be human. Use human language. Keep it in the same general tone as your site, plus a tablespoon of helpfulness.Also: Look decent. Use a nice font and take a little time to format the message so it matches your site’s look and feel.
Are you paying attention to me?	Reassure visitors that you’ve noticed this issue and are working to fix it. Ideally, of course, this is actually true - you have a solid error-reporting system in place and someone taking care of problems.”We’ve noted this problem and are working hard to fix it.”
What should I do now?	Be specific about what action the visitor should take next, and make sure there’s a way for them to do it.One of my pet peeves is seeing a classic “…please contact the webmaster” message with no email link provided.”Try changing your search and try again.”
What if that doesn’t work?	Provide an alternate path the visitor can take to achieve his or her goal, even if it means displaying (gasp!) a telephone number or email address. I know, it’s extreme.”If you still experience problems, we’re here to help. Call us toll-free at 1-800-xxx-xxxx and we’ll get things straightened out for you right away.”

Error Message Examples

Example #1 - Microsoft

I get this annoying message every time I forget and try to grab Windows updates using Firefox (my preferred browser).

Things to notice:

1. No indication an error has happened. I suppose it’s not a real error then.
2. No apology. Not too surprising.
3. It looks like something out of a 1995 Perl script - plain black text on a white background. Not that I have any problem with Perl, mind you.
4. A very matter-of-fact tone. There is no attempt to humanize the message, or add any personality.
5. A plus: quite clear directions
6. Another plus: three separate action paths visitors may follow (get IE, use the Download Center link, or turn on Automatic Updates)

Example #2 - Google

I received this error message a week or so ago, while Gmail was down for a couple hours.

Things to notice:

1. This is identified as an error message, with the additional flourish of being a “Temporary” error. Immediately reassured. Somewhat, anyway.
2. Some formatting and design are present. It looks nice.
3. An apology - “We’re sorry”
4. Not blaming the visitor - “your Gmail account is currently experiencing errors.”
5. Explaining what’s happening, with just enough detail to be reassuring.
6. A “What next” - “try accessing your account again in a few minutes”
7. Several alternate, but unsatisfactory actions -Try Again, and Sign Out. I’ve resigned myself to having no support for free Google products, so am not surprised there’s no way to get more help. Bummer, but offset by the freeness of Gmail.

Oops! We’re out of time!

Sorry, I’ve run on so long that we’ve reached the end of this blog post. For more reading enjoyment, please visit our other web usability or conversion optimization posts.

If that doesn’t work, leave questions or topic suggestions in the comments and I’ll get right back to you.

 

I’m sure most of you know that Google Analytics (GA) Content reports display pageviews and unique pageviews, along with other metrics.

Did you know, however, that you can create and track virtual pageviews? You can create a virtual pageview anywhere that you can call a javascript method. This virtual pageview will show up everywhere in the Content reports section that a ‘real’ pageview will.

This means you can track clicks on Flash movies, different CSS layers, pdfs, images, docs…the list goes on.

I am going to give three examples of how you can use virtual pageviews to help increase and track conversions, but first I’ll demonstrate how to set them up.

How to Implement Virtual Pageviews

This example assumes you are using the new Google ga.js code and that your normal GA tracking code is on your page. Here’s an example of that tracking code for reference:

<script type="text/javascript">
var gaJsHost = (("https:" == document.location.protocol) ? "https://ssl." : "http://www.");
document.write(unescape("%3Cscript src='" + gaJsHost + "google-analytics.com/ga.js' type='text/javascript'%3E%3C/script%3E"));
</script>
<script type="text/javascript">
var pageTracker = _gat._getTracker("UA-youraccountcode-1");
pageTracker._trackPageview();
</script>

To create a virtual pageview, you have to call the following javascript method.

pageTracker._trackPageview('/The-Name-Of-Your-VIRTUAL-Page');

Note: The Urchin method is different: urchinTracker('The-Name-Of-Your-VIRTUAL-Page');

Ok, I realize that’s a bit cryptic. I will give you an example, and then explain further.

Let’s say you have an image of a red rose, and when you click on it, a larger version of the image opens in a CSS layer:

<a href="path-to-open-large-red-rose-css-layer" title="Red Rose"><img src="path-to-my-red-rose-thumbnail" /></a>

All you have to do to create a virtual pageview (which of course means you can then track how many people clicked on the red rose image) is to add in the above javascript method and make up a descriptive page name:

<a href="larger-red-rose-image-css-layer-path" title="Red Rose" onclick=”pageTracker._trackPageview(’/LargeRedRose’);”><img src=”path-to-my-red-rose-thumbnail” /></a>

Note: The page name is virtual. You can call it anything you want, and it will show up in GA as a normal pageview. You can also include virtual directories. I could have used:

onclick="pageTracker._trackPageview('/Flowers/LargeRedRose');"

It’s up to you how you name the virtual pageview, but plan it out well so you can easily find and group things in the GA reports.

You can create as many virtual pageviews on a single URI as needed. If you have 50 flowers on a single web page, you can create 50 different virtual pageviews.

(If you’re wondering if there are other ways to call the method, the answer is ‘yes.’ You can use any technique you want, including unobtrusive javascript, as long as the javascript method is called.)

Below are three examples of how you might use virtual pageviews.

1. Downloadable White Paper

Let’s say you are a software development company, and one of your web site conversion goals is for visitors to download your white paper on “Smart Ruby on Rails.”

The code on your web page to download the PDF white paper is:

<a href="whitepapers/smart-ruby-on-rails.pdf">Download White Paper Now</a>

To create a virtual pageview for the download, insert the javascript method I showed you previously so it is called when a visitor clicks on the download link:

<a href="whitepapers/smart-ruby-on-rails.pdf" onclick=”pageTracker._trackPageview(’/WhitePapers/SmartRubyOnRails’);”>Download White Paper Now</a>

Now every time a visitor clicks on that link, it will be recorded in GA as a pageview. It will show up next to all of your other pageviews in GA as /WhitePapers/SmartRubyOnRails. (Don’t worry. My next example will show screen shots from GA that demonstrate this.)

Now you can track how many people download your white paper (convert). To go even further, you could define this new virtual pageview as a conversion goal in GA to get even more information, but I’ll get to that.

2. Measure Content’s Effectiveness

On our Search Engine Marketing page, we have four small case study graphics. If a visitor clicks on any of them, a new CSS layer appears with more information about that particular case study.

We wanted to track which graphics visitors clicked on the most, or more specifically, if there were one or two that received a significant majority of the clicks. This would give us an idea of what content is more compelling to our visitors.

We achieved this by creating different virtual pageviews for each case study graphic. Here is code for two of the graphics:

<a href="javascript:appearNow('f50');" onmouseover="document.fort.src=fortover.src" onmouseout="document.fort.src='img/fort-50.gif'" onclick=”pageTracker._trackPageview-> (’/caseStudyViews/Fortune50′);”><img src=”img/fort-50.gif” name=”fort” /></a>

<a href="javascript:appearNow('tech');" onmouseover="document.technical.src=techover.src" onmouseout="document.technical.src='img/tech.gif'" onclick=”pageTracker._trackPageview-> (’/caseStudyViews/HighTech’);”><img src=”img/tech.gif” name=”technical” /></a>

You can see below that the virtual pageviews are included with the ‘actual’ pageviews in the reports:

Because our virtual pageviews all start with /caseStudyViews/ we can easily find them in the reports by performing a search on ‘casestudy’:

Over time we found that more visitors clicked on graphics that had larger numbers, like “1500%”.

Of course you might argue that you could use the Site Overlay feature in GA to see the same thing, but GA’s Site Overlay is pretty unreliable. You could use a tool like CrazyEgg or ClickTracks, but then you’re going to yet another solution, AND by using virtual pageviews you can actually test whether or not content helps drive conversions.

3. Measure Whether or Not Content Drives Conversions

Let’s look at techsmith.com, and specifically their SnagIt product offering for this next example. (Note: The following is a fictitious example.)

techsmith.com wants their web site visitors to download a free trial of SnagIt in the hopes that they will upgrade to the pay version. So, a download of the free trial is a conversion.

techsmith.com has various content pieces like movies, demos, and images in their assets arsenal, and they want to find out which content drives the most conversions.

In this specific example, they are testing a movie and want to track whether or not a site visitor watched the movie before downloading the free trial.

To accomplish this:

1. Create a virtual pageview when a visitor clicks the Play button:

<a href="javascript:playVideo();" onclick=”pageTracker._trackPageview(’/FreeDownloadTest/PlayVideo’);”><img src=”myPlayButton.jpg” /></a>

2. Create a Goal in GA. The Goal URL will be the Downloading page.

This is the URL the visitor is taken to after they have clicked to download the free trial. Remember, this is the end goal…conversion point.

3. Define a Required Funnel Step.

Now you have defined your conversion point for your new Goal, but you want to track if the visitor watched the video before converting. To track this you need to put the name of the virtual pageview (that is created when a visitor hits Play) into Step 1 of the Goal Funnel, and make it Required.

The Funnel Visualization report will only show a conversion if the visitor triggered the virtual pageview (clicked the Play button) and then downloaded the free trial at some point after that.

Every download of the free trial will count as a conversion. Now you can see how many of the visitors that converted watched the video before converting.

Of course, clicking the Play button does not ensure they watched the video, but if you could call the javascript method at some point from within the movie…this leads to an:

Idea: Another interesting use of this technique might be to have a Flash demo make a javascript call to create a virtual pageview somewhere within the demo. This way you could tell if a visitor actually went through the entire demo etc. You can probably think of other ways to incorporate this as well.

Other Ideas?

Well, I’ve suggested three uses of virtual pageviews to help you increase and track conversions. Feel free to share your suggestions.

This is a follow-up to a post I wrote last April - Why the Google / DoubleClick Deal is Brilliant - when Google announced they were buying DoubleClick. I made a number of predictions at the time and thought it would be interesting to see how I fared. Turns out I did pretty well overall, although the jury’s still out on a few predictions.

I’ve given myself a grade for each of these predictions. Self-grading seems unfair, so let me know if you disagree with my grading scheme.

Prediction #1

SNAPSHOT:
• Search revenues are projected to triple to $44.5B over the next 5 years (Piper Jaffray). Google is already positioned to capture a significant majority of those revenues.
• But Search only represents ~50% of all online advertising revenue (current & projected).
• This deal is about positioning Google to go after the other 50% of revenue projected to come from non-Search ads.

Note: The projections for search revenues have actually increased.

PROJECTED GO-TO MARKET STRATEGY:
Google is most likely going to bolt the DoubleClick ad server platform and ad network on to their existing AdWords system over time. By doing so, Google will open up a lot of additional graphic/rich media/video ad inventory for AdWords, addressing a big hole in their current offering. While they’ve had the ability to serve up banners and video ads for a while, non-text ads currently account for only a tiny slice of their overall inventory and revenue. When this happens, Google will not only be serving those ads – they’ll be selling that inventory. Why? Because Google’s huge network of 500,000 advertisers will help publishers monetize that inventory better than they can themselves.

MY GRADE: A
Nailed this one. The Google content/placement network has expanded to cover the DoubleClick network, and the ad targeting capabilities have been enhanced to include some additional DoubleClick features. From the Official Google Blog: New enhancements on the Google content network.

Prediction #2

And if Google needs to partially subsidize some of those high-value publishers by paying guaranteed CPM’s for some time, they can do that. They certainly have the cash. And that will quickly become a moot point as the big brand advertisers (and their agencies) who are currently buying that inventory directly at high CPM’s will realize they’ll need to buy that inventory via Google or risk losing it to a higher bidder.

MY GRADE: TBD
I jumped the gun on this one, as there’s no evidence that Google is out there buying market share. But it could still happen at some point. They’ve certainly outbid Yahoo to serve search listings on other sites (see their CNN deal, for an example)

Prediction #3

They will probably continue to offer the DoubleClick ad server product as a standalone product for quite some time so as not to alienate existing advertisers/agencies or publishers, but the real goal will be to make participation in their marketplace a no-brainer for everyone by offering better distribution and accountability for advertisers and better monetization for publishers. As Google gains more advertisers and publishers, the holdouts will have a harder time resisting the gravitational pull.

MY GRADE: B
DART, DoubleClick’s ad server product, is still going strong – and probably will for as long as it’s more profitable for Google to offer it. Which is likely years, especially for large advertisers and agencies.

I’ll retroactively give myself an A if/when Google stops trying to promote DART in favor of the AdWords platform at all.

Prediction #4

WHAT IT MEANS FOR:
Advertisers - Win
Online advertising gets more efficient. With Google as the primary hub and marketplace, advertisers get to deal with one entity and cut one check for most of their online advertising. Managing those campaigns will be very complex, but will still be much more straightforward than dealing with individual publishers with different pricing and deal structures.

MY GRADE: B+
Yep. More and more of our clients are looking to traffic their banner campaigns via Google, and Google continues to gain an increasing share of advertisers’ budgets.

Prediction #5

WHAT IT MEANS FOR:
Publishers - Win
The enhanced AdSense program becomes even more attractive by offering better monetization via advertisers who are willing to bid up effective CPM’s for their inventory.

The market has spoken on this one. More inventory is coming online for Google all the time, as measured by impressions.

Probable side effect of this: Many mid-tier publishers who are currently fielding their own ad sales forces will likely decide they can make more by just selling their inventory via Google, meaning a lot of ad salespeople will likely get pink slips.

MY GRADE: B
Are some of the jobs being lost in the economy coming from the ranks of online advertising salespeople? I haven’t seen any hard evidence one way or the other.

Prediction #6

WHAT IT MEANS FOR:
Other Ad Servers - Lose
Over time Google will have the only ad server that will seamlessly integrate the dominant Search platform and Graphic/Rich Media/Video ad serving and reporting – thus partially resolving one of the major headaches advertisers and agencies face. And they’ll probably offer it at a substantially lower price. Some specialization will be required to compete.

MY GRADE: A
GoogleClick now has 77% share in the ad server market, as seen here on attributor.com’s post. While smaller ad networks continue to proliferate and benefit from the growing online advertising pie, the chances of anyone making much market share headway are slim.

Prediction #7

WHAT IT MEANS FOR:
Performics – Lose
Google will have to spin out or sell Performics in order to avoid the perceived conflict of interest inherent in a Search Engine owning a Search agency. The uncertainty accompanying this speculation and the eventual transition will likely result in at least a moderate loss of new and existing clients for Performics in the short term. (Full disclosure: My firm competes with Performics). Google has already updated their official statement on the deal relative to Performics from “no plans to dispose of it at this time” to “evaluating all strategic alternatives for this business.” So the writing on the wall is pretty clear at this point.

MY GRADE: A
GOALLLLLL!!!!! Google Sells Off Performics to Publicis Groupe, announced August 6, 2008. This one had me nervous for a while, as there was no word from Google for a long time, despite pressure from Danny Sullivan – “Open Letter to Google: Do the Right Thing, Divest Yourself of Performics” and, to a much lesser degree, from me – Google’s Stake in Search Engine Optimization. But then Google finally responded with an announcement in April, 2008 – Selling Performics Search Marketing

Phew.

My grade would’ve been an A+ if it hadn’t taken a year for Google to even say they intended to sell Performics.

Prediction #8

WHAT IT MEANS FOR:
Other Search Engines & Portals – Lose
With Google locking up so much inventory on 3rd party sites, other engines and portals will have an increasingly hard time growing their ad revenues beyond the market growth rate and will continue to lag behind Google in both revenue and prestige.

MY GRADE: A
Witness the 77% number above – and the Microhoo debacle.

Prediction #9

WHAT IT MEANS FOR:
Agencies - Win
As mentioned above, manually configuring and integrating reporting from Search and non-Search advertising is a major headache and productivity drain for most agencies. Google will go a long way toward easing this problem by seamlessly integrating Search and non-Search ad serving and reporting into a single platform.
Even more compelling for agencies, though, will be the opportunity to boost profits by cutting their ad server bills by switching over to Google’s lower cost ad server.

MY GRADE: C-
I missed the boat on this one. Certainly search agencies are taking full advantage of Google’s dominant market position and growing suite of valuable (and free) marketing tools. But many agencies are wary of Google’s power and, more to the point, Google’s intentions. Many fear that Google is trying to obviate the need for agencies by making it so easy for advertisers to work directly with Google to run their campaigns.

Personally, I’m not one to throw my body on the tracks of progress. I love that Google is bringing transparency to the advertising world. But by making it seem too easy, Google risks making it seem like there is no specialized knowledge required to get the most out of their platform, which of course isn’t true. Just because I CAN remodel my own kitchen doesn’t mean I should, as my wife can attest. Google also hasn’t helped their cause with agencies by occasionally trying to go direct with clients – effectively pushing out an existing agency. That doesn’t play well with agencies generally, but especially those for whom Search is a new tactic.

But I’m getting off-topic here…

Anyhow, most agencies haven’t embraced Google’s rise as a good thing as I’d predicted. But they are increasingly working together, out of necessity if nothing else.

Prediction #10

WHAT IT MEANS FOR:
Google – Big Win
The media has missed the point on this deal by viewing it as a defensive move or a technology play (which is really laughable given that Google already has the world’s largest ad server and marketplace). While they certainly paid more than they would have otherwise in order to outbid Microsoft, it’s important to point out that Google doesn’t think or operate with a defensive mindset. Virtually every move they’ve made has been part of a larger strategy. This is no different.

With this move, Google has put themselves in a position to capture an even bigger slice of the global online advertising revenue pie by giving them direct relationships with the advertisers/agencies and publishers who will be responsible for deciding how those funds are spent.

Google already accounts for 19% of all online advertising revenue (32% in the US) – almost all of that from Search. With global online advertising projected to grow to $81.1B by 2011, it is not unreasonable to think Google could break even on this deal in as little as a few years IF they execute this product integration and transition well.

Given their deft execution to date and penchant for blowing away expectations, I wouldn’t bet against them. This deal positions them to dominate the non-Search online advertising market in the same way that they’ve come to dominate the Search market – by making it easier and more efficient for advertisers/agencies and publishers.

Don’t be surprised if 50%+ of all online advertising revenues go to Google in five years (and possibly a decent chunk of offline ad revenues – but that’ll have to be a topic for another post). If that scenario plays out, this could be looked back on as one of the best deals in the history of business.

MY GRADE: A
OK, the whole “best deal in the history of business” thing may have been a bit overstated (I was on a roll by that point). But there’s no arguing that this deal has been a huge win for Google, and arguably the main reason Microsoft felt so compelled to pursue Yahoo – as they realized (too late?) that was their only hope of creating a contender to the Google juggernaut.

Prediction #11

DON’T BE DISTRACTED:
The press will play up these issues related to the deal because it’s easy to generate buzz on these topics, but they are ultimately insignificant.

Antitrust Concerns
There is very low likelihood that the FTC will block this deal, as the online advertising market is so difficult to define and, more importantly, there is no reasonable argument to be made for how consumers, advertisers or publishers will be harmed as a result of this transaction. And Google’s success with all of this isn’t assured anyhow, despite my views. The FTC can’t block this deal because Google might execute brilliantly.

Privacy Concerns
People will make a lot of noise about privacy concerns, but in the end most will act in their own economic best interests and ignore the conspiracy theories about what Google might do with all the data at their disposal (regardless of how much truth there may be to them). And while Google has certainly used the massive amounts of data at their disposal to benefit themselves by improving their products, there is no indication that they have done so at the expense of their users. In fact Google has a pretty strong record on protecting privacy rights, as Matt Cutts explains here.

MY GRADE: B+
I hadn’t expected Congress to get involved in investigating the deal. That worried me for a bit. I guess I should’ve expected a holdup as a result of the EU review. But the deal ultimately went through with few problems.

MY OVERALL GRADE: A-

(Why yes, I am using my own proprietary algorithm to weight the grading, thank you for asking)

I still contend that Google’s future growth potential is huge in the offline advertising space, but I’ll cover that in an upcoming post. Hint: People are freaked out about Google controlling half of the online advertising space, but online accounts for only ~10% of all media dollars. Imagine if they can make a dent in the other 90%. They’re closer than you think…

I’ll be off for a long overdue vacation by the time you read this, so apologies for the lack of responsiveness to comments, etc.

Once the party’s over, are you leaving customers out in the cold?

Thank you pages typically don’t receive much design or marketing attention. After all, by the time a site visitor sees a Thank You page the chase is over, right? The visitor purchased a widget, filled out the signup form, or downloaded a white paper - in other words, the web site has won, and another conversion stat has been chalked up in the company’s analytics package. Success! Check out our positive ROI!

It’s understandable, really. Most companies have much larger issues to worry about - such as growing their sales and trying to read Google’s mind (or Cuil’s) - than the design and messaging of their Thank You pages. As a result, most Thank You pages are utilitarian afterthoughts that leave customers cold. They become merely a convenient way to track conversions and deliver boring stuff like receipts and download links to the customer. Oh, and maybe some unemotional boilerplate from the company’s “Team” thanking the customer for their transaction. If you’re lucky.

Thinking outside the conversion

As the web matures, however, more organizations are mastering the basics of doing business online. Their shopping carts work. Their landing pages are motivating. Their site content rocks. And everything’s on an iterative testing schedule to squeeze out incremental conversion gains.

With the basics out of the way, and companies looking to solidify their brand, maintain their credibility, and generate those highly valuable long-term customer relationships, more are paying attention to the smaller touch points. They’re making sure that every interaction, even those not directly related to gaining a sale or conversion, receives close scrutiny and marketing optimization efforts.

From this perspective, no touch point is insignificant. Each represents an opportunity.

A quick example of touch points

I’ll use my recent clothes-shopping trip to Nordstrom as a quick example. Notice the sheer variety and number of touch points involved in my purchase. Each had some effect on my perception of the brand and willingness to interact further with Nordstrom.

TOUCH POINT	SCORE	BRAND INTERACTION
Legacy score	6	My current brand perception of Nordstrom, based on past interactions.
Web site	+4	Checked Nordstrom.com for sales and specials, and right on the home page saw a promo for an upcoming in-store sale - 20% off!
Brick-and-mortar store parking	-3	Drove to the mall on the appointed Sale day. Struggled with parking outside Nordstrom, cursed the parking lot design while repeatedly seeing the Nordstrom sign on the building.
In the store	+2	Store’s interior design is classy and clean, not too crowded, clothes neatly displayed. (Observed to self in totally unrelated aside that men must not be willing to trundle up to the 2nd floor to shop, since their department is always on the ground floor, right by the door. Women, on the other hand, are evidently shopping masochists).
In the Women’s Department	+3	Emerged into a confusing array of choices (-2) but was approached very politely by a helpful sales clerk who guided me to the Sale area (+2) and then (this is important to me) left me alone (+3).
Caught between sales clerks	-6	Gathered items to try on, then (ironically) couldn’t find the sales clerk I’d been so happy to dispense with earlier. Found another to let me into the locked changing stall. When I emerged with my selections, my original sales clerk approached me, and a turf war ensued between her and the second clerk while I stood by and became increasingly irritated.
At the purchase counter	+1	Once sparring clerks worked out their sales commission split, the purchase went smoothly and professionally. I needed one blouse in a smaller size, and the clerk ordered it for me from another branch store, promising it’d be delivered to me within a few days.
Purchase completion	+1	My clerk thanked me with a brilliant smile and called me by name. I received my items neatly folded into branded tissue paper and carefully placed in a sturdy, attractive carry-bag (also branded). Ever felt a Nordstrom receipt? No skimping on paper there. It’s rich, thicker than the average receipt paper, and conveys a tactile reminder of quality.
Shipped item arrival	+2	Received my smaller blouse a mere two days later, in a padded envelope shipped via UPS. The outer package wasn’t branded, which gave me a moment’s pause. But the blouse was folded and wrapped in branded tissue, and arrived promptly and with a minimum of wrinkles. This is two days after I completed the purchase, and yet I’m being reminded of the brand in a way consistent with my in-store experience.
NEW SCORE:	10	Overall, my recent interactions with the company were positive, which improved my perception of the brand. Given the funds, I’ll probably buy from Nordstrom again in the future.

What do your visitors see after they click Submit?

An online Thank You page is basically the equivalent of the Nordstrom “Purchase completion” touch point I listed in the above example. Should it be the first area on a site to optimize? Of course not. But once you have the basics working, there’s opportunity here for cross-selling, brand reinforcement, clarifying the next steps in a larger transaction, and what my colleague Lance Loveday calls “goodwill enhancement”.

In other words, ignoring your Thank You pages is a missed opportunity.

Two good Thank You page examples

#1: CrazyEgg

CrazyEgg’s a pretty cool click density visualization tool we’re trying out as an alternative to a company whose name will go unmentioned due to poor customer service (touch points, again - only they ended up with a negative score). This screenshot is of the Thank You page seen immediately after signing up for a basic account.

Positive things to notice here:

• Clear branding
• Clear, very legible statement telling me what just happened.
• A minimum of text. I can skim & understand the page quickly.
• Clear statement telling me what will happen next (an email will be sent)
• What I see is identified (”Your account details are displayed to the right”) - although arguably that should be self-evident.
• An obvious action for me to take next (click the “back to dashboard” button)

A couple of minor negatives:

• No actual “thank you” - jeez, I just gave them 9 whole U.S. dollars, you’d think they could at least say thanks.
• No personalization. By now CrazyEgg knows my name, but chose not to use it here.
• The account summary titled “Basic” could have a clearer heading.
• The “Back to Dashboard” button is slightly un-intuitive for a first-time visitor. What IS the dashboard? Was I ever AT it before? Couldn’t say.

#2: Musician’s Friend

Musician’s Friend is, well, exactly that. This screenshot was taken right after I purchased a lovely electric-acoustic guitar and mini-amp to try out.

Positive things to notice here:

• Clear branding
• Clear, highly visible statement telling me what’s happened (Your order is complete).
• My name. It’s reassuring, and despite knowing exactly how it’s done behind the scenes, I still feel special.
• A thank you!
• Clear statement of what happens next (a copy will be sent to my email address)
• An order reference #
• Clear contact info - phone # and live chat button
• Security reassurances - a brief statement in the top area, a detailed Secure Shopping Guarantee at the bottom, and the Verisign logo
• Clear call to action - “Continue Shopping…”
• Cross-promotions - Clear, categorized links invite me back into the site. I can’t be positive whether my guitar purchase triggered the Guitar category displaying at the top, but that’d be very appropriate.

A couple small nits:

• The page could be better optimized for cross-sales, using the current order data. I was offered complementary items such as guitar picks and stands while I shopped, but this is another opportunity to make a gentle offer. In this case, I’d actually forgotten a necessary amp cable and some extra strings. I would’ve appreciated seeing some order-specific offers.
• There’s no receipt! I’m used to printing a receipt after a purchase, even though I expect a receipt copy via email. While I appreciate the clean page layout, I’d still like a link to print a record of the purchased items, price, etc.

Thank You Page Optimization Guidelines

An optimized Thank You page contains answers. Answers to typical questions such as:

VISITOR QUESTION	OPTIMIZATION GUIDELINE
Is this the right place?	This is where consistent design and branding come in. The Thank You page should be consistent with the look and feel of the checkout process.
Am I done yet?	Provide some indication that the process - or at least this portion of it - is complete. For example, “Payment has been received,” “Your inquiry has been sent,” etc.
Remind me - what did I just do?	For some transactions, a record of what just took place is very helpful. For purchases, a receipt. After a long or complex registration form, it doesn’t hurt to be specific in the process completion statement - i.e. “Your CAR LOAN application has been submitted.” Remember, you may not have 100% of your visitor’s attention.
Are you grateful for my business?	The words “Thank you” are such a minor thing, but so humanizing.
What happens next (if anything)?	Be specific about what will happen next. For example, an email will be sent (to what address?), or someone will call them (who? when? why?). For transactions with a significant offline component, such as loan applications, provide a clear, informative list of steps the applicant should expect to happen next.
Is there anything I SHOULD do right now?	Be specific about any immediate actions your visitor should take. For example, “Wait! You’re not quite done! Click the DOWNLOAD link to complete your purchase.” or “Check your email for a confirmation link. You must click this link in order to activate your account.”
What else?	IF the transaction is completely finished (i.e. an ecommerce purchase), consider providing cross-selling or other opportunities to interact with your company. NOT, however, when this might distract your visitor from additional steps they need to complete.
What if I have questions?	Provide at least one way to contact you. This goes a long ways toward reassuring the visitor.

Your reading of this blog post is complete!

Thank you for reading this Closed Loop Marketing blog post about optimizing Thank You pages! Don’t worry, you don’t have to do anything else right away - your brain will absorb and process this information all on its own.

Next steps:

• Optimize your own organization’s Thank You pages
• Pass this post on to a friend
• Contribute your thoughts - leave a comment

Questions? Leave a question in the comments, and I’ll be sure to get back to you.

With great predictability, Google’s recent announcement about improved Flash indexing was rapidly followed by numerous client requests for clarification:

“What does this mean for SEO?”

“Do we still have to include HTML text on our Flash pages?”

“What do I need to do differently?”

In response to these and other requests, we’ve compiled the following summary and recommendations to assist our clients (and blog readers) in determining the best course of action relative to Flash, Google and SEO.

Google’s Improved Ability to Index Flash – In a Nutshell:

• Adobe recently developed an optimized Flash player for search engines to use and is collaborating with Google and Yahoo on implementation.
• This means Google can now explore Flash more like a website visitor does.
• This new technology only covers text and links, which is a small part of most Flash content.

Recent Updates Regarding JavaScript:

• Initially, Flash triggered by JavaScript was not going to be indexed because search engine robots typically do not execute JavaScript code (Google’s June 30th announcement).
• However, a recent update to the official Google announcement (circa mid-July) indicates an update has been rolled out to enable support for common JavaScript techniques for embedding Flash, including SWFObject and SWFObject2.
• This means that Flash triggered by JavaScript now has the potential to be crawled.

What Has NOT Changed:

• Images, videos, animations and graphic text in Flash are still NOT indexable.
• This change also does not yet take into account any Meta data or formatting markup in the Flash file.

Flash and SEO Best Practices are STILL Recommended:

• Even considering this announcement, SEO best practices prevail because the amount of text on a page (or in a Flash file) still contributes heavily to the relevance and value search engines ascribe to a page or file.
• In fact, the amount and type of relevant, indexable content on a page is even more important now. The greater the competition, the greater the need for relevance.
• It is still a best practice to follow the “tried and true” Top 3 SEO Strategies for ensuring pages containing Flash are search-engine friendly and accessible (see the “Additional Information” below for more detail).
• These best practice recommendations are based on the fundamental difference between being indexed and ranking well in search results.
• Indexation does not mean improved rankings. A really snazzy site with very little indexable content, is still a really snazzy site with little indexable content, regardless of whether the Flash component can be crawled.

What Developers Should Do Differently with Flash Files:

• Separate URLs for each set of content within the Flash experience is still the best way to ensure content gets indexed (see the “Additional Information” below).
• Separate URLs also contribute to increased link popularity and the opportunity to create optimized Title Tags and Meta Descriptions for each page.
• Because search engines are able to see text within Flash files, any text that is not important to search results (e.g. copyright information, “loading” messages) should be replaced by images to avoid indexation.
• Google’s response to questions regarding the concern over duplicate content is:

Serving the same content in Flash and an alternate HTML version could cause us to find duplicate content. This won’t cause a penalty — we don’t lower a site in ranking because of duplicate content. Be aware, though, that search results will most likely only show one version, not both.

Timing for Implementation:

• Google has acknowledged that the impact of this improved capability will take time to fully propagate through search results.
• There is no word on timing for Yahoo implementation, but Adobe is working with Yahoo on next steps.
• Adobe has not yet made the technology available to MSN, possibly as a result of MSN’s production of the competing Silverlight product.

ADDITIONAL INFORMATION ABOUT FLASH AND SEO:

A) Top 3 SEO Strategies for Flash:

1. Same Page HTML: The simplest search-friendly approach.

Include important text and links from the Flash movie elsewhere on the same page, as plain HTML.

2. Progressive Enhancement: The best future-proofing approach

Using a CSS ‘DIV’ tag, the page developer would define a certain area of the page for this display. Then, using JavaScript, test for whether the Flash plugin is present and (by default) whether JavaScript is enabled. If they are, the Flash code is written to the page with JavaScript.

With this approach, if the visitor does not have either JavaScript or the Flash plugin (like search engines do not), plain HTML will display. Otherwise, the Flash will display. The idea is to make the page friendly to all types of devices. This approach can be extended to encompass text readers, mobile devices, etc., helping to build flexibility into your page coding approach.

Potential drawbacks: possible layout issues with old browsers when the text is very long.

See Bobby van der Sluis’ article for Adobe on developing Flash sites using progressive enhancement for more information.

3. CSS Layers: The third best approach.

This uses CSS to define two layers, one above the other, positioned in the same space on the page. The Flash element is placed in the top layer, while the text and links are placed in the lower layer, hidden by the Flash.

Potential drawbacks: possible display issues with older browsers; does not add the flexibility of a full progressive enhancement approach; a slight chance that this may be viewed as an attempt to ’spam’ search engines - however, using this method in a conservative manner is unlikely to cause any penalty.

B) The importance of having separate URLs for each phase of the Flash experience:

Excerpt below taken from the Search Engine Land post by Vanessa Fox:

…[T]he searcher experience is better served by Flash implementations that provide a unique URL for each set of content. Some Flash implementations dynamically load text as the user interacts with the application, but the URL remains the same. In this scenario, Googlebot can now follow those interactions (in a limited way) and if the URL doesn’t change, then all content that is dynamically loaded as the interactions progress is associated with a single URL.

…This means that if the content that is dynamically loaded into the Flash application from the fifth interaction matches a searcher query, that Flash application may be served in the search results. But when the searcher clicks over to that result, the content won’t be found on the page. The searcher will have to interact with the application until that content is loaded. Searchers may instead feel frustrated and abandon the page.

For the best user experience and higher conversion rates from search, Flash developers should be careful to avoid this situation by creating distinct URLs for each piece of content. This implementation helps the Flash site be more viral as well, as users can email, Digg, and otherwise share the content more easily.

Google acknowledges this scenario may not be an ideal searcher experience, but points out that other non-HTML file formats such as PDFs have the same limitations. When a searcher clicks through the Google search results to a PDF file, the content that matched the query may not be on the first page of that PDF and the searcher has to scroll through the file to find the desired content.

C) Further Reading about Google, Flash and SEO:

1. Official Google Webmaster Central Blog June 30 Announcement (now includes an update to the original announcement)
2. Summary and explanation of announcement from Search Engine Land by Vanessa Fox
3. Andy Beal’s post on Marketing Pilgrim cautioning against all-out Flash adoption
4. Erick Schonfeld’s post on TechCrunch includes a good point: “Becoming visible is one thing, actually ranking highly is another”
5. Adobe Developers resource web site – clearly covers the use of the “Progressive Enhancement” approach

As I’m sure many of you already know, Google made a much-appreciated change to their Keyword Tool last week. Now, Google’s keyword tool provides actual search volume numbers, as opposed to the green bars you’ve seen in the past that represented low, average or high search volume.

One of our first questions upon hearing this news, was whether the search volume reported out by Google was based on the exact keyword or based on the volume for the keyword plus all of its Phrase and Broad Match iterations.

As it turns out, the tool allows you to pull search volume for all 3 of Google’s match types. For example, we entered ‘kids cell phones’ into the tool and searched it 3 separate times, each time changing the pull-down for match types. The data Google provided:

What a great feature! 

Also important to note, is that the stats reported are based on the number of searches performed in Google and the Google Search Network only. Therefore, the 4,400 approximate searches you see above for the exact keyword ‘kids cell phones’ does not include the number of searches performed in other engines such as Yahoo or MSN.

The ability to pull search volume by match type is a unique feature that makes a very helpful addition to Google’s already-robust suite of tools for search engine marketers!

Screen resolution.

No, I’m not talking about how Hi-Def your HDTV is. (Although I’m sure you’d be happy to tell me.)

Yes, I’m talking about your online audience’s display resolution. While the subject sounds much more stale than your HDTV, it just might help afford you an upgrade.

Don’t Design Outside of the Box

Many designers seem to design without giving any thought to the visible width and height of the end users’ display. Having dabbled in design myself, I have also been guilty as charged, many times over.

As technology advances, enabling screen resolutions to grow and costs to drop, the problem of designing sites too large to fit users’ screens seems to have lessened, but it is still a problem. As your users’ screens get larger, guess what? So does your designer’s, and probably at an even greater rate.

In fact, right now I am staring at a screen that takes up half of my desktop and has a resolution of 1920 x 1200. I’m pretty confident that mine is bigger than yours, and unfortunately there’s a decent chance my design will reflect that.

Of course if you are a true purist you will say that one should always design using a fluid layout, like Jakob Nielson points out. While I completely agree in theory, in practice, designers are too meticulous about their layouts to let them be stretched, squashed, poked and prodded at.

What Screen Resolution Should I Design For?

Note: The following statistics were taken from W3CSchools.

Display Resolution Statistics:

Yes, many more people are using high resolution monitors, but the majority of people are viewing web sites at 1024 x 768 resolution.

So, design for 1024 x 768, and remember the browser window takes up a lot of that real estate.

Personally I try to use 960 for the width. I attended a few Cameron Moll sessions last year at WDW Seattle, and his argument for using 960 was a good one. It’s divisible by 3, 4, 5, 6, 8, 10, 12, 15, and 16 which means your grid options are endless, and of course it fits within the most popular 1024 width.

That’s all fine and good, and isn’t too shocking, but what is unfortunate is how many designers ignore the arguably even more important height dimension.

If you need to keep content above the fold, I would use 525 as your line of demarcation. This is especially important for landing pages where the call-to-action needs to be ‘in your face.’

Yes, You Should Give a Damn About Screen Resolution

Look at this example.

Using 1024 x 768 for my display resolution, I performed a search on Google for ‘kids toys’ and clicked on this recognizable brand in the Sponsored Links. This is what I saw above the fold:

No, the giant in-your-face area is not clickable.

If you were to scroll down you’d see a couple of smaller call-to-action areas that are, but you’ve already lost me.

First of all, that whole area should be clickable, but minimally the primary call-to-action NEEDS to be above the fold. If the designer had tested this on the most common screen resolution, I think they would have made some different decisions.

In contrast I did a search for Luxury Cars and every one I checked out worked well on my screen:

I would change a few things on this landing page, but the point is the designer got all of the pertinent information above the fold.

I will concede that it’s much safer now to design web pages and landing pages bigger, but look at the statistics, decide where your target audience most likely falls, and then be sure to test your designs before launch.

OK, a little background: Sandra and I did a webinar for Aquent last month, and had over 2000 people register to attend. We thought that was pretty great. They also mentioned they were going to also put our slides up on SlideShare, but we didn’t think too much about that at the time.

In a follow-up post from the Aquent blog, Matt Grant noted that SlideShare had featured our presentation on the home page at one point, and that over 8000 people had viewed it so far. Wow. As Sandra said, we’d have spent more time on those slides if we knew they were going to get that kind of play. Anyhow, I went over to SlideShare to check it out and halfheartedly thought “I wonder if we’re on the Most Viewed list.” But we weren’t… Wait, that only shows the current day’s most viewed. What if I click This Week? Bingo:

That’s 8595 views for our presentation vs. 7354 views for Ms. Jolie.

As it’s taken me two weeks to complete this post due to a blog platform issue, my fifteen minutes of fame are long gone now. But still… to have our presentation get more views than something titled “Angelina Jolie Pictures” feels like a pretty major accomplishment. Especially with such unalluring title slide.

And now we’re the 6th Most Viewed for the month, still just ahead of Angelina. See it for yourself, before we roll off.

Maybe there’s a market for this whole Web Design for ROI thing after all…

While at SXSW in March, I ran into Shannon McKarney from Jaded Pixel, the company behind the popular Shopify hosted e-commerce platform. Shannon asked if she could interview me for the Jaded Pixel blog and I agreed. I’m cross-posting the interview below. The original is here.

What inspired you to write the book?
We were struck by how many of the sites we worked with suffered from the same design and usability problems. But that was really a symptom of the way people think about and manage their sites. So in addition to providing some concrete design guidelines, we wanted to try to change the way people think about web design. Personally, I wanted more people to see the money they’re leaving on the table by not designing their sites well.

How the Book Was Born
I had just finished a presentation on Designing for Conversion at Web Design World in Seattle in July 2006 and was answering some audience questions in front of the stage afterward when a kind but serious-looking gentleman handed me his card and said “I really enjoyed your presentation. How’d you like to write a book about it?” A quick glance at his card identified him as Michael Nolan, Sr. Acquisitions Editor at Peachpit Press/New Riders. After a stunned silence while my mind fixated on the long-held perception that ‘I don’t write so good’, I smiled crookedly at him and said with all the gravitas and certainty I could muster “Ummmmm… Sure?”

After that inauspicious start I managed to write a book proposal, get it approved by the editorial team at New Riders, and built out a schedule and outline. I was on my way. Then it hit me: “Writing a book is hard. There’s no way I’m going to be able to write a book while managing a business, keeping clients satisfied enough to continue paying me, and being a decent husband and father to my 2-year old son. And I’m certainly not going to get it done on my own before my twin-pregnant wife delivers seven months from now.” Fortunately, the obvious solution to my problem was sitting in the next office.

I’ve been fortunate to have Sandra Niehaus, our VP User Experience and Creative Director here at Closed Loop Marketing, as a friend and colleague for the last three years. In that time, she’s bailed me out of more jams than I can count. And when I proposed the idea of coauthoring this book together, she did it yet again. In retrospect, I can’t imagine trying to write a book like this without Sandra. She brings a designer’s eye and real-world Web design and programming experience that is crucial to conveying so many of the concepts in this book. She also art-directed the whole thing, no small thing for a graphic-rich book about design. In a nutshell, anything good in this book was probably Sandra’s idea. I take full responsibility for everything else.

You have a great approach in the book - you talk about what you’ve seen; you say what will help people with these issues; then you say “stop reading and try this!” and give examples to really bring it home to the reader. Where did this approach come from and what’s the feedback been on it?
I wish I could say our approach was the result of a grand vision we had from the start. But the reality is that we iterated on it for months. We wanted the book to reflect the design philosophy we advocate, so it seemed appropriate to have a strong call to action component. In trying to come up with a name for that section, one of us said “The effect we’re going for this is to have people stop reading and try whatever it is we’re recommending.” So that’s how we came up with that heading. Incidentally, the hardest part of writing the book was figuring out how to structure the content – just like designing a good site.

What were some of the biggest website design “fails” that you’ve seen in your career? (naming no names, of course)
My favorite one is when we were reviewing a client’s shopping cart and got to step 3 of their 4-step checkout process only to find no way to get to step 4. They’d inadvertently commented out the ‘Continue’ button a few days earlier, so it was impossible for someone to actually buy anything on the site. And they wondered why sales were down…

You’ve had a very diverse background before you got to the web space. How do you think that influences the way you look at web design?
I’m not a designer or a programmer, so I don’t see the same things that most Web professionals see when I look at a site; I only experience the interface itself. In that way I think I’m more representative of most Web users. But my background in economics compels me to think through the implications that the user experience has on the web metrics and the business metrics of the organization behind the site I’m reviewing. So I sort of reverse engineer the business model of every site I see and think of ways to tweak the site to help them improve their metrics.

What’s next on the horizon for you?
I’m most focused now on growing my company. We do a lot of Search Engine Marketing and Conversion Optimization work, which is a lot of fun and fulfills our goal of helping organizations get more business benefit out of their sites. On the publishing front, we’re starting to play with video. My sense is that some well-produced multimedia videos would do an even better job at conveying the principles we recommend than 2-D static book format. I’m also speaking at various conferences, which I really enjoy.

Have you seen any Shopify stores that exemplify good design?
I just bought a T-shirt from the Tesla Motors site because I can’t afford one of their cars just yet. There was one unique element of the checkout process that really stood out as a best pract