Should Brands Pull Back on Detailed Analytics?


There is tons of data that can be gleaned from detailed analytics, but do brands really need all of that data? Does it really benefit them in the long run?

Digital analytics tools have become more powerful. But that doesn’t mean you need all that power, especially if you’re a brand and not an e-commerce provider.

First, let’s take a look at the superior capabilities of analytics tools.

Today, instead of dropping a single line of code on the page and hoping for the best, today’s analytics implementations are paragons of multi-channel complexity. The rapidly maturing field of tag management now permits dropping one of their lines of code on the page – but at least in one common format, instead of hoping for the best, they allow you to put a dozen different rules and data gathering tags into a “container” that executes every time a page is loaded.

Add to that much-vaunted “predictive modeling” – essentially a way of retargeting visitors with more relevant ads — and you have the beginnings of a closed-loop architecture that allows you to market with more accuracy and with better ROI than ever.

Today’s tool kit may also include multivariate “optimization” tools, and lately, even QA automation tools.

But what if you’re not trying to “convert” visitors in the classic sense? What if you just want them to interact with your content (to sell ads) or interact with your brand to develop loyalty that plays out at the mall?

Chances are, you don’t need to know all that much about the visitor, and trying to find out more than is needed starts to drain away ROI. Chiefly, you want to know a few key metrics — and these are relatively easy to obtain.

Brands will want to know how many visitors came, how many came back again, how many pages they looked at per visit, which pages were most popular, and whether they shared the content. Yes — you can retarget, too. If you stream video, you will also want to know how many video starts you had; and how much of the video the visitors saw before leaving. If you have apps, you will want to know how many got downloaded, how many got opened, and how many got deleted. In-app measurement is usually unique to the app and broadly defined measures are much harder to define.

Brands today can skinny-down their analytics demands and still get excellent value out of the exercise.

But wait: because of CMS complexity and developer inattention, the cost of knowing anything at all can be almost as much as the cost of knowing more than you can use.

Long gone are the days when Web pages were simple affairs being served to users like fancy text files. Now the content is aggregated from multiple sources, and no page actually exists in full until it is executed in the browser. Client-side tracking makes this a bit easier but conflicts arise. Conflicts in tagging can result in gross under- or over-counting, and that is why extensive testing becomes a critical part of analytics for the enterprise.

Brands can certainly know less and prosper. But they also need to plan deployments efficiently, train their developers, and enforce uniformity across properties and channels. Simplicity in the dashboard ought to be backed by clarity in deployment.