MAKING ALL THE NUMBERS MEAN SOMETHING

4th Jun, 2015
by Todd Wackrow

Cta Join Us

Sign Up

Get Union Blog
updates straight
to your inbox

Sign Up

×

In a digital world where you can measure every metric you can possibly think of, there is a real risk of it all becoming meaningless. Digital agencies (us included) can sometimes be guilty of assaulting our Clients with an array of statistics, but without true context or purpose this can often just be data overload, with it impossible to decipher what it all means.

As a business obsessed with delivering ROI, to try and counter “data overload” syndrome we have reinvented our approach to Insights & Analytics, with an aim to actually use less stats rather than more. This starts with digging deeply into what good actually looks like, because unless you have a true purpose to what you are trying to achieve stats are of no use anyway.

"The starting point is just being clear on why you are here" 

The starting point is just being clear on why you are here. What I mean by this is that businesses or brands (in most cases at least) exist to sell something, so any investment you make should be helping you sell more of said something. Whilst this seems a no brainer, it’s scary how many businesses have a website or Facebook page because “everyone else does” and don’t truly understand the role it plays for their core business purpose which is (generally) selling something to somebody else for more money than it cost you to produce it.

Once we understand what good looks like, we then work out what digital metrics can be used to help monitor whether our efforts are being successful. From here we have also started building customised 1 page Dashboards which collate all key data, without the risk of it being lost in the clutter. Whilst Google Analytics is an incredible piece of technology, and we still use it to supply our key stats, the “can’t see the wood for the trees” (whatever the hell that means) factor is high.

Our final approach to data is also trying to never use a Stat unless we are also including the following:

1. Context

An engagement rate of 2.03928%... great…. Stats have no meaning by themselves so we always try to give a statistic some context through either comparing it to Last Years result, industry standards, competitor results, etc. Even just comparing it against the previous month helps you understand whether you are trending in the right direction.

2. Insights

Aka a plain English translation on what this Stat actually means.

3. Recommended Actions

So what? We force ourselves to ask this question every time we feel like rolling out a Stat, and to ensure we accompany it with a recommended action to improve performance and ROI.