This is what you see:

This is what you see when you click the feedback button

A visitor will have pressed the feedback button because they have something on their mind.   What I like about this is that the first thing they see is what is on the mind of all the other people that have submitted feedback. A list, prioritised by a simple voting system, shows the visitor what the biggest gripes are.

If you see your own gripe there, then vote for it.  If not, lodge your feedback and it’ll be there for others to vote on when they leave their feedback.

The next nice thing about it is that if the website developers/owners decide to implement an idea that someone has left, they can indicate that the work is planned, in progress or even completed.  Instant feedback to the visitor – letting them know that yes, something is being done about it – great – we’re being listened to.

So – it provides you with great information that helps you create a development roadmap for your website – ever wondered what shall we do next to our website? Well you’ll get tons of inspiration from this.

Finally, a small word of warning, although your website should cater for the needs of your visitors – make sure that it is the segment of visitors that you want to appeal to that you are developing for.  It won’t always be possible to cater for everybody – this is where your judgement comes in and your knowledge of the goals of your website.

Custom Reports became a bit of a poor relation.  When it first came out I couldn’t think of a good reason to create a custom report – so I left it well alone.  After a while of using Advanced Segments I was pleased with myself as I had worked out how to follow a tagged email campaign link through to any transactions associated with it (and therefore revenue) and the products that were purchased.  This was great – the ability to see exactly what items were purchased after somebody had clicked on a specific email link in a campaign.

It was a tedious process though, you had to identify the link in the campaign that had generated revenue (taunted by the not-hyperlinked ’15 transactions’ it was associated with), then create an Advanced Segment for that link (it was tagged using utm_term, which is presented in the ‘keyword’ dimension), then apply that segment and look at the Ecommerce transactions to see which transactions (and therefore products) were associated with that link.  Long winded huh?  Especially if you wanted to investigate several revenue-generating links.

Custom Reports to the rescue!  Using Custom Reports, it is simple to set up a report that uses ‘Campaign’ as the main dimension, then drills down to the ‘keyword’ dimension, followed by ‘transaction’, followed by ‘product’.  Drop the metrics you want to see along the top, and you’ve got a drill down report that begins by listing your campaigns, and allows drill down all the way to product level.  Fantastic.

Easy to create a drill down report

No more laborious Advanced Segments to create!

Previously, using this variable completely broke the bounce rate reporting as it counted the usage of _setVar as an ‘interaction’.  A bounce happens when the visitor only has 1 interaction with the site, so even if they only visited 1 page but the _setVar was used, it wouldn’t be counted as a bounce because pageview+setVar = 2 interactions.  The following image shows what happened to us the last time we tried to use the _setVar method…

_setVar destroyed our bounce rate metric!

To cut a long blog-comment-ping-pong game short, I drilled down to a single fact/assumption that underpinned his analysis which I’m not 100% sure about.   The assertion is most easily explained with an example:

1. A visit to the website occurs 30 days ago from a ppc campaign “randomCampaign”.  You can set up an advanced segment to constrain your reports by “Campaign=randomCampaign” so you can see visits to the website from this campaign.  All good stuff.
2. Now imagine that the same visitor decides to visit your website again, but this time as he knows the url he types it into his browser and for this visit is therefore “direct”.
3. In a couple of days we run a report on Google Analytics constrained by the “randomCampaign” segment for a time-span of say, 5 days.

Underpinning the methodology Avinash is using, he is expecting that the report in step 3. includes the visit that occurred in step 2., despite that visit being a “direct” visit and not a “randomCampaign” visit.

My understanding is that the segmentation will act on the visits that occurred in that time-period and the dimension that is being segmented by must apply to those visits, not previous visits outside the date range.

Now we don’t seem to agree on this, but luckily it is something we can test.  So starting today, I’ll begin an experiment to test the above and I’ll report my findings here in a week or so…

==================================

Couple of days later…

So here is the quick test I ran:

1. I cleared all the cookies from my browser.
2. I visited my website from a referral: gatest.somewhere.com.  I made sure I knew which page I visited.
3. The next day, I went to the website again, but this time I typed the URL directly.

My original thinking was that visit in step 2. would be a source=referral, and the visit in step 3. is a source=direct.  Wrong wrong wrong!!

Despite the visit in step 3. being a ‘direct’ visit, GA is clearly reporting the source for both visits as the original source that is associated with me.  This explains why when you set up an Advanced Segment, say for this example where ‘source = gatest.somewhere.com’, both visits in steps 2 and 3 will appear in your segmented reports.  Which validates Avinash (surprise surprise) and forces me to accept a whole new paradigm…

So in summary: when you view the ‘direct’ report in GA, the visits that are registered may or may not have been ‘direct’ – only their very original visit to the website was ‘direct’…  This applies to any of the source dimensions – which poses several interesting questions…

Edit: further clarification here at Justin’s Epikone blog on how GA tracks bookmark visits.

Edit (29/09/11): They released it yesterday: and according to a couple of sources, it’s going to cost about $150k per year. 

But I sincerely hope that they are working on one – I’ve mentioned before that it’ll be great news if Google release a version of Google Analytics that you have to pay for.

So here I’m beginning to compile a list of the things that I come across every day that I wish is in Google Analytics, and would be happy to pay for:

In no particular order…

• Advanced Segmentation: I would love for AS to be enabled for Goal Funnels…
• “This report is based on sampled data.  Learn more.”  I would really like to not see this message.  It means that the numbers I’m looking at are not the real numbers.  Even if it took a little longer for the report to run, I’d like the ability to see the complete picture and not their “sampled data”.
• Data backup.   One day it might happen – someone with administrator access on my account accidentally (or maliciously – eek!) deletes my GA account.  Disaster!!!
• Retrospective filters.  You apply a filter and it acts on your data retrospectively Implemented with Advanced Segmentation.  Thanks!
• Multiple dimensions.  At the moment you can set a ‘user defined’ variable, but you cannot add several custom dimensions that you can use to slice and dice your data
• Annotation.  Like the charts you see on Google Finance or Google Trends, I’d like to be able to annotate my charts with significant events that I know of that might explain certain trends etc.  Done!
• More AdWords integration.  Sometimes there are campaigns and AdGroups that you see in analytics that confuse you – which ad is that again?  To be able to click through to see the ad text and other details would be nice
• On charts overlay profiles (or segments as they might be called) – so if you have multiple profiles set up for a single website with each profile having a different filter applied (you might have one that shows organic search only and another with paid search only), I’d like to be able to take any chart and add data to it from another profile.  So you’d get a data series for each profile you added. Implemented with Advanced Segmentation
• Related to the previous one – it would help enormously if you could duplicate a profile from an existing profile – that way you don’t have to set up your goals etc all over again… Implemented with Advanced Segmentation
• Ecommerce transactions.  These are a must for any website that sells stuff online.  When you look at a particular report, you can click the Ecommerce tab and see the revenue associated with that referral source, or region etc.  It even tells you the number of transactions that are associated with that record.  What I’d *love* to be able to do is then click on the ‘transactions’ value, to arrive at the list of those transactions so I can identify the products purchased.  I will then be able to quickly ask the question: “what do people that come from X tend to buy the most”.  Instead what I have to do now is set up an Advanced Segment constraining the report by X in order to see what the transactions were, and whilst this works, it is a very time-consuming process.

We have implemented a URL validation step when we process a response
to make sure that when people call a page they use the correct URL.
If they use an incorrect URL, then they are sent a 301 redirect with
the correct URL.

The URL in our sitemap is in the format:
http://www.domain.com/index.html?whatever=value

We’ve now had errors showing up in Webmaster Tools, with it saying that Googlebot is coming across too many redirects in our sitemap URLs.  The problem with Googlebot is that even though we put the correct URL in the sitemap, it doesn’t use that URL to make the request – it omits the index.html bit, contracting it down to:
http://www.domain.com/?whatever=value

So our server sees this ‘incorrect’ URL, issues a 301 with the
‘correct’ URL (that has the index.html bit in it), but then Googlebot
doesn’t follow that URL faithfully and again tries to request the URL
without index.html in the path.  So our server again issues a 301
redirect, with the correct URL and here we go off on our infinite
loop.

So no wonder we get the error message:
URLs not followed…. [sitemap] contained too many redirects.

I think this is a bug as the 301 redirect clearly sends the redirect
URL, if Googlebot followed this redirect URL faithfully then we
wouldn’t see this issue.

Here is the sitemap error in more detail (substituted our actual domain for a pretend one).

HTTP Error:
Found: 301 (Moved permanently)

http://www.domain.com/?param=whatever1
http://www.domain.com/?param=whatever2
http://www.domain.com/?param=whatever3
http://www.domain.com/?param=whatever4
http://www.domain.com/?param=whatever5
Jul 20, 2008

Double checking the sitemap file, these URLs are in the right format complete with index.html.

Why does Googlebot strip out index.html?

However, I would estimate that the vast majority of web teams do not have the resources to manage ongoing testing alongside all the usability, redesign and fixing of their site.  The clever e-business knows that their tech/analytics talent is valuable to them, but few pour resources into an area that is hard to calculate a clear ROI for.  So, when opportunities for the web/tech/ecommerce teams come along that put them in the spotlight and show their worth, they need to be taken.  So how is web-team-public-relations related to multivariate testing?

Done properly, multivariate and A/B testing will significantly increase conversion rates – ideally conversion in a process that involves revenue.  If you can show that your team have increased the ‘add to cart’ or the ‘checkout’ conversion by 10% – that makes a good start for justification for more resource.  But wait – we don’t have that resource yet.  So this is where we get to do some PR and get test results at the same time.

Set up your test framework in whatever system you like – Omniture Test & Target is good if you have the cash but Google’s Website Optimizer comes for free.

Then after you have identified an area to focus your tests on, announce to the rest of the company that you are running a competition to see who can come up with the best content/design for that area and that you will be testing the 6 best entries on the live website.  Getting people from other departments is great – you get a fresh insight that is more layman, and probably closer to that of your customer/visitor than your web team.  Hopefully they come up with good ideas, some of which you take forward into the test.

The nice thing about this from a PR point of view is that it has several clear stages with a specific overall purpose.  The purpose is to increase revenue, something that the whole company should easily buy into, the thing being tinkered with (the website) is easy to understand and relate to, and the process fits a competition structure well with opportunity for regular communications with the rest of the company on progress.  The end result should be a formula for a better conversion rate (and thus revenue) that has had lots of internal publicity – great well done to the people that came up with the winning test version and well done web team for coming up with this idea.

Exec: “You want more resource to do more of the same next year?  Sure!”

A typical A/B or multivariate test might take a page portion and then serve up several static variations. Sometimes static variations aren’t good enough though. Most eCommerce websites are database driven and use templates for product pages that are populated with information specific to that product. The template knows what product info to load in because the page might be accessed via a URL with identifiers in the query string: e.g. http://somesite.com/product.html?productid=1234

This product page knows that it has to load up the details for product 1234.

When you want to start doing more complex A/B tests, where the data for your variations also comes from the database, you have a slight problem in that the alternative content for the test is managed in the website optimizer interface – how do you get dynamic content out of your database for your test variations?

To get around this, you can use Ajax to grab the dynamic content relevant to that particular product page, and use the Website Optimizer to simply modify parameters in the Ajax call. This might be implemented by creating four server-side functions that are accessed by Ajax, each returning a variation on the original test content.

In Website Optimizer, when you declare which part of the page you are testing, rather than wrapping the content section, wrap the piece of Javascript that sets which function the Ajax request will call (or Javascript that sets an Ajax parameter):

<script>utmx_section(“AjaxSection”)</script>
<script>aj_fn = “variation1″;</script>
</noscript>

Then, when you proceed through the experiment designer to add new variations, just add:

<script>aj_fn = “variation2″;</script>

Where “variation2″ is the name of the function the Ajax will call to return the “variation2″ content.

Alternatively, as mentioned previously, instead of creating a function per content variation just alter an Ajax parameter so that the function returns different content.

This combination of Website Optimizer and Ajax makes for an extremely powerful technique. It’s pretty easy to implement too.

Personally, I would be over the moon if this were to happen, because then we would be able to request features with more of an expectation that they will take them seriously (not that they don’t now, it’s just that if we paid for it then they would have to take us even *more* seriously).

One of the good things about GA is that they keep your analytics data for a very long time. We’ve had our account with them since 2006, and being able to go back that far to analyse traffic and behaviour is very powerful. Sometimes though, it would be nice to be able to delete or ignore some data – for instance one particular institute in Tempe, US, decided to build a bot that executes javascript and then crawl all over our site. For the most part, we can happily use GA in the knowledge that most spiders don’t execute javascript, but this javascript-executing-bot now appears in my GA data (as GA data-collection is javascript driven).

So I’ve got this nasty spike of data that I’d just like to be able to select, then hit the ‘ignore forever’ button.

I guess, that when Google do decide to tap into the thousands of organisations that really want more features and are happy to pay a premium, this would be one of the many features I’d ask for… as well as more Goals, better page-flow analysis, page-rendering-time data, more than one custom dimension, the ability to break out traffic from Google across the country-specific domains, etc etc etc…

You can do reporting to different levels depending on whether you want to measure the total response time, the perceived render time etc. Because the data comes from every single visit to your website from real website visitors (rather than an automated datacenter-based script) we get ‘real’ data on performance. Unfortunately they only keep data for 33 days, I’m working on them to increase that

Here’s a chart that shows the performance of our product pages over the last couple of weeks, segmented by region:

The good thing here is that each datapoint is the aggregate of all page visits during that time period, not just a single request from a monitoring station. I know now that this is what our end-users experience – a constant source of debate when management are experiencing (or told about) something different to what you are reporting…

http://www.gomez.com