August 28, 2013 ↘︎

Measure people, not traffic.

I recently attended the Adobe Digital Marketing Symposium in Sydney and what a sellout that was.  800 people registered; 1,100 rocked up, way to go!  It was a great day with lots of interesting customer success stories. 

As always we get excited by the direction that Adobe are taking the general product suite. But one thing continues to strike me – 90% of companies are still measuring traffic, not people, despite the capability of all of the platforms nowadays.

We need to stop worrying so much about Visits and Page Views and all of the other base metrics – because when taken in isolation, they don’t really help you all that much to improve the user experience, or to improve your conversions.

Measure people.

It’s people that come across your digital channels; it’s people that purchase or sign up for something.  Wouldn’t you be better off knowing more about the people engaging with your digital channels, rather than looking at more traditional segments such as traffic sources, campaigns etc ?

Looking at people means that you need to be able to view cross-visit activity…looking at a visitor rather than a visit.  It means you need to be able to build “visitor” level segments.  And all of the platforms can do that now.

More often than not, you’ll find that people don’t convert on the first visit.  They might be researching your service or product, and then comparing it to others.  It might take 3 or 4 visits before they actually convert, but their behaviours across those non-converting visits are definitely important factors that you should be looking at.

Having all of those cross-visit metrics available allows you to begin to better profile your converting audiences – to figure out common traits that they have.  It’s far more insightful knowing what content they browsed over different visits prior to converting; knowing the average length of engagement each time; knowing how they got back to you etc etc.

It’s even better if you’re creating them an account when they sign up, or if they log-in to purchase, because you can associate not only their new ID across seemingly anonymous visits, but you can also load customer demographic data and look at behaviours by demographics.

Using profiling techniques, you can better understand how different age ranges, genders, household addresses and so forth, engage with your brand.  Not to say that you can’t profile with traditional web data, you can to a certain degree, it just becomes far richer when you start to augment the data.

Now you’d be able to understand how Females differ to Males, by Age Group, when they come in from an email campaign, or through organic search…  How long does it take them to convert?  How much are they spending?  How many products do they buy?  Do they convert through another channel?  Do they engage through different devices, or multiple devices? 

A basic example:

If we were purely looking at the source of the last visit to better understand performance, we’d have something like the following:



It tells us a fair amount, but it’s in the aggregate. 

We’re looking at traffic, not people. 

We’re not seeing how they breakdown in terms of demographics, or other internal segments.

By looking at visitors, then breaking them down by the source channel, we get a better understanding.  For example, it appears that as our customer base ages, they typically look at more pages, but our younger audience tends to spend a bit more per person.

We can also see fairly significant differences across revenue by different channels and age ranges.


From a profiling standpoint, you’d want to start comparing these guys to each other, and to those that don’t convert, or engage more.  Profiling should start to show the main differences between them.  It’ll start to highlight opportunities for you, such as new content, possible usability fixes needed, search terms etc.

So if we only look at traffic, we miss a large part of the picture.

We need to be looking at our customers, our visitors.  We need to be looking at people, rather than traffic, if we are really to get value out of digital analytics.

Even if we don’t have access to the demographic data, looking at visitors, not visits, helps form a much better understanding of behaviour than individual, disconnected visits.

So come on folks, stop worrying so much about Visits, Page Views and the likes, and start considering your Visitors.

DB logo
DB logo
DB logo