eCommerce Analytics

Over the last few weeks Google has been rolling out remarketing features to users of its AdWords platform. These new features allow companies to continuously target visitors with ads after they’ve left its website in the hopes of luring them back. Think email remarketing but with ads on a content network.

Remarketing (often called retargeting) is not new. Dozens of companies have been offering this functionality for some time, including Criteo, Fetchback, Dotomi, and Adroll.

So why all the fuss? One big reason. Nearly every online retailer already has an AdWords account. This means the friction of setting up a remarketing program has suddenly been drastically reduced.

The Case For Google Remarketing

For many online retailers, the easiest way to dip a toe into remarketing is to do it with Google’s new features. There are no contracts to sign, no new relationships to forge, and a familiar PPC cost structure. It’s also relatively easy to set up a simple remarketing campaign that retargets site visitors with a generic ad.

Most important, the reach of Google’s AdSense network cannot be beat. No matter where your visitors go after they leave your site, you can be sure that they’ll soon be seeing an ad from Google’s network.

The Case Against Google Remarketing

The biggest drawback to using Google for your remarketing program (if you’re a typical online retailer) is the sheer scale of Google’s business. Google has to make its features appeal to every company in the world in order to make the tiniest contribution to their overall growth. They simply cannot afford to concentrate on adding features specific to integrating with e-commerce processes. Instead, Google’s business dictates that they must provide a self-service set of tools in the hopes that companies will do the heavy lifting with regards to business logic.

The result of Google’s reliance on self-service tools is that you’ll likely need a front-end developer to modify Google’s tracking code on the fly and a lot of patience for manually defining groups and rules in the AdWords interface. And if you want to integrate your remarketing program and reporting with your other systems, it’ll be up to you to do it since Google offers no professional services.

Should You be Using AdWords Remarketing?

If you are a tech savvy organization or are looking to get started with a basic remarketing program, then Google Adwords could be the perfect solution. It has familiar interface, a name you more or less trust, and solid functionality.

But if you are looking to implement a more sophisticated remarketing program or you need some expert help to guide you then you will want to look elsewhere. Like the rest of its tools, Google’s remarketing features are made exclusively for the self-service crowd.

Do larger eCommerce retailers have more loyal customers than small retailers? Our data suggests maybe.

We dug through web traffic metrics for 520 retailers to see how many times the average visitor returned every month. Retailers with over 2 million unique monthly visitors have 67% more sessions per visitor every month than those with less traffic volume.

One possible interpretation is straightforward: bigger retailer = better brand = more loyalty. Another interpretation is that visitors return to bigger retailers more often simply because big merchants have a large selection of goods where customers can meet a variety of their shopping needs.

These theories are not mutually exclusive, of course. But are there other explanations that we’re missing?

We’ve wrapped up our first look at web analytics market share among online retailers and if you’ve been thinking about ditching your current package for the free Google Analytics, it appears you’re not alone. Google Analytics is now in use by 44% of online retailers with Omniture close behind at 34%. Coremetrics rounds out the big three with 17% market share.

As expected, Google Analytics dominates with smaller e-commerce players while Omniture owns the 5 million uniques/month and above segment. Coremetrics makes a respectable showing in the 500,000+ monthly uniques segment.

33% of mid-sized retailers use two or more analytics packages — nearly double the rate of larger retailers.

Of those, nearly all are using Google Analytics with either Omniture or Coremetrics. Is this a sign that retailers are kicking the tires on Google Analytics or is it just an inexpensive way to validate the figures from the more powerful packages?

Update: The full e-commerce web analytics market share report is now available for download in PDF format.

Fun fact from a study we’re cooking up: almost 2 in 5 retailers who are using Omniture or Coremetrics for their web analytics are also using Google Analytics.

Are they using Google Analytics to verify the numbers from their industrial strength packages or are they contemplating a switch? It certainly makes sense to use a second package if you want some assurance that the numbers you’re seeing are legit, especially if that package is free. But it’s not a stretch to imagine that these retailers might be giving Google Analytics a test drive.

Bonus fun fact: over 1% of online retailers are using both Omniture and Coremetrics. That’s some serious analytics firepower. And while you may have heard of these retailers, they aren’t the household names you might think.

Stay tuned for more web analytics analytics.

I was looking through a Coremetrics white paper today entitled Optimizing Your Marketing Mix in a Down Economy and was struck by the sheer amount of website interaction data the Coremetrics platform tracks:

• Every web page viewed by visitors

• Specific paths that visitors take through key site processes

• Web page point of entry, navigation path, and departure path taken by visitors

• Every banner ad, email campaign, affiliate link, search engine keyword (paid and organic), blog, news article, and any other source that brings visitors to the web site

• Every product, room, flight, or merchandise item that visitors click on, view, or interact with, and reserve, book, buy, or abandon

• Every newsletter signup, customer registration, and opt-in identification action taken by visitors indicating that they wish to be contacted

• Every important attribute of the visitor’s browser, including screen resolution, plug-ins, time zone, language, IP address, and domain name

Read Our Next Challenge: Turning a Firehose of Data into a Trickle? »

I’ve somehow ended up with a copy of Men’s Health in my house as a result of my holiday travels. It’s been a few years since I last picked up a copy and now having checked out the most recent issue I can confidently say that Men’s Health stinks at statistics.

Read Which X-Axis is the Right One? »

Omniture’s sales and marketing department are simply world class.  Not only are they beating the pants off of CoreMetrics, but they’ve sold a vision of complete information awareness that online retailers are eagerly buying into.  Almost every single one of our customers has bought Omniture or is about to.  And every single one of them has asked the same question.  Now what?

Read Congratulations! You’ve installed Omniture! Now What? »

The knock on B2B data mining has always been that there isn’t B2C-like data available. Instead of multiple transactions that give us customer behavior patterns, we have company demographic information (industry, company size, revenue), some information about the person from the company who we’ll deal with (position/title), and where that person came from (lead source). It’s not behavioral data, which we know to be inherently better as a predictor than demographic data. But some data is better than none, right?

And we can certainly create transactional data that gives us some behavior pattern. If we throw in the contact schedule - the touches - from your company’s representatives, don’t you have a transactional pattern of both buying and non-buying customers? Coupled with the demographic data, you can drum up a model that predicts how many touches a lead might need to become a client and maybe a best guess at the path that should be pursued with a new lead.

More to the point of this post, this is the great thing about click-through data: it has a transactional quality. In fact, it just might be the transactional data for B2B companies. (Aside: This is also one of the reasons why companies like Omniture are becoming so notable: they provide some behavioral patterns, however small.) If we can combine click-through patterns from the person representing the prospect company with the company’s demographic information, then we might have a real interesting model that determines just how serious a lead is about buying from you and their company’s relative experience level with your product area.

Let me close out this post by refuting two of the main complaints about B2B data and its unsuitability for data mining-based models.

There’s Not Enough Data

Everybody loves data mining when it comes to consumer-focused companies. The vast amounts of transactional data are transfixing. The thinking goes something like this: “I’ve got hundreds of thousands of transactions here so whatever our predictive model spits out must be right.” Well, this may be true. And it mayn’t. But that doesn’t make a model built with less data any less compelling. It just means that one model has more data points. Don’t feel inadequate for the difference. Just make sure that you have data that’s important to the business problem you’re trying to solve. For example, if you want to know the next-best product for newly-minted customers, then you’d better have a solid set of second-time customers who bought a bunch of different products. Do you need thousands of these second-time customers? C’mon.

Missing and Bad Data

Isn’t this a reality everywhere? Even consumer-focused companies (with hundreds of thousands of transactions) have this issue. Oh, and I have a suggestion on what to do with that missing and bad data. Throw it out. Chances are, it will have absolutely no effect on the predictive models, unless of course all of the missing or bad data has a common characteristic that isn’t found in the rest of the data. For example, let’s say you’re building a model that predicts the next software product that a first-time customer might want from your company. Well, if everybody that bought a specific product as their first purchase is missing a zip code, then you can’t very well throw all of those records out. It would skew the model irreparably. But as long as the missing data is evenly distributed throughout the records, don’t be afraid to trash ‘em.