Recently, I found myself needing to describe – even defend – web analytics to people who had in the past dabbled with the numbers, without making any real headway.

I can’t say I blame them. Google analytics is, let’s face it, not great at reflecting the situation on the ground when it comes to business performance (It’s not just GA but we’ll stick with that for obvious reasons). As most people know, it’s a trend based system, based on anonymous and often incomplete bits of data, extrapolated from when the tracking script is allowed to pass information back to Google’s vast data centres. Sometimes the numbers are bewildering – it can tell you your site is bursting with so many visits that you should advise the board to acquire a call centre and fulfilment operation, yet in reality, yesterday’s sales amounted to one confirmed, two maybes and four unsolicited job requests.

So why the disconnect?

A contributing factor to the problem is often how the we perceive the value – or intent – of a visitor. Business owners have a habit of imagining their website as a real bricks and mortar shop, located in an exclusive part of town. People journey from afar, arrive at the grand entrance and begin pacing methodically around, staring intently at all the wares on display. A nod from the assistant on every corner, and a transaction must surely follow?

This misconception about online behaviour, coupled with an onslaught of data, is no doubt responsible for a lack of faith in Google Analytics.

My analogy of the website is perhaps a little disheartening at first. Your website is basically a tent at the world’s largest outdoor market, surrounded by other tents almost identical to yours (I say a tent because it is a temporary, less substantial structure, and that is representative of the trust people put on websites during online shopping compared to the real world physical retail experience).

Vast hoards of people are constantly in flux, swarming around, in and through your tent. Many poke their head through a flap and quickly back out. Yet more step through various entrances and blindly run through, some not even there to shop! A lot of people are clearly just seeing what the noise is about. Amid the chaos, money does sometimes appear in the pot, but you didn’t get chance to see who threw it in there, and where they came from. In other words – pure chaos!

How do analytics segments help?

Imagine being able to make those random people disappear, revealing the serious shoppers walking carefully around your tent. For now, get rid of those just scurrying through. And imagine that the side flaps are rolled up so you can see exactly where these serious shoppers have come from. Segments provides this clarity.

You can set segments up to examine groups of people that interacted with 2 or more product pages on the site – Browsing with intent! Segments for time spent on site – Researchers! Segments for those that accessed a specifications tab, or availability calendar – Planners!

Couple segments with metrics such as time between visits, and attribution reporting, and you can glean real insights into the consideration periods linked to your products and services. Identifying customer profiles based on data can be a powerful driver for your marketing plans, and with segments, Analytics really helps flesh out these pictures.

Once you start using segments, you drill through the noise and find out who your best customers – potential and existing – are, where they came from and what they have in common. In fact, without segments, Analytics is just sampling noise, and really lacks any useful accuracy. At some point, we have all ended up in those dreaded discussions around generalised, un-segmented analytics data, resulting in wild assumptions and theories. You really do not want your long term marketing plans based on shaky foundations.
Segmented data linked to results (be it sales or higher levels of engagement) gives you the grounding to make the right decisions. Yes, the volumes can be small, but the insights are much richer. And I argue that, rather than being seen as an ‘advanced level’ analytics skill, even newcomers should get to grips with segments straight away.

Matt Davidson

Contact Matt on +44(0)1392 275554 or  email: matt (at) goldladder.co.uk