Five frustrations about measurement and how to resolve them
Whilst working in digital marketing I saw first-hand the time that it takes to produce weekly and monthly reports for clients. In fact, I got so frustrated with this laborious task that I decided to create software that would automate the entire data collection and presentation process.
It became a personal challenge of mine to take on the data frustrations that myself and others had and tackle them head on. It means that I’ve become a bit of an expert on all the little annoyances of analytics and enjoy having a good old moan about them from time to time. Read on to hear my personal gripes in the world of measurement and how to turn these into, what I like to call, Data Joy!
1) Meaningless Data.
Having come from a science background into marketing I think I’m possibly overly aware of this. Other than costs, revenue and profit you have to look quite hard to find the data that is really, truly important. The majority of data is much more inconsequential. Part of what we do at Trackpal is help people to focus on the important metrics and cut out all the pointless data.
My advice: Reports should include only what is useful, and not more. Only report on those things that reflect your top-level strategic objectives. They shouldn’t include made up metrics that try and simplify things but only confuse people, or where the definition of the metric might change over time.
Whatever you do, do not jump on them. These are often created by dominant players in the market that everyone feels like they have to ‘game’ in order to beat. For example, a Facebook ‘Like’ becomes meaningless when there are hundreds of fake Likes. It’s the same with Google and the emergence of link farms.
In the world of data there is always a new metric or insight popping up or an entirely new data source altogether. A lot of these are really interesting to monitor but some are just pointless and short lived. My advice for this frustration is not to focus on short-term metrics as measurable KPIs unless you have a good reason to do so. (i.e don’t try to reach 10,000 Facebook Likes just for the hell of it.)
My advice: Targeted metric measurement is more important than measuring the current ‘big thing’ in analytics without a good reason.
3) Measuring too much and reacting too quickly.
I love analytics but I know that a good marketer does not need a live stream of numbers to best market a product. It’s one piece of the pie. When I started Trackpal analytics solutions were all about dashboards and instantaneous views, which can be useful but it can also be extremely distracting. Sometimes going with your gut is better than what the numbers tell you from the outset. What I try to do is make it easy for people to have the information they need so they can look at trends over time and use the information to make informed decisions.
My advice: Focus on what you know you *need*, not what you think you may need. You can always get that later, *if* you need it.
4) Big Data.
Yes, Big Data is important. Big Data has also been around for a long time though so it’s nothing new. It just seems like it’s another bandwagon that people are panicking about and therefore desperately putting in place a Big Data strategy that they might not actually need. You might have a Big Data issue but often when people talk about Big Data – especially in a digital marketing context – the thing they’re talking about isn’t really Big Data at all. I’ve written about the Big Data buzz in more detail over on our blog.
My advice: Allow yourself pause for thought before you decide to dive into absolutely all the data at your disposal.
5) Confusing Labels.
One of my personal frustrations is when a new data set comes about with a confusing label that conflicts with other similar names across other sources. People often do this when reporting. For example, Impression vs. Reach, Traffic vs. Visits. Data should be tidy and consistent. By all means come up with your own labels but choose these carefully and stick to them.
By way of comparison, how screwed up would scientists be if the definition of a metre varied as much as an “authority score”, or sentiment, over the course of a century, let alone month to month?
My advice: Create a ‘Data Glossary’ for internal use where you can specify what you are measuring. Or use ‘Datamorpher’ to collate your data into a manageable format.
So those are my personal frustrations within analytics and my advice on how to avoid them. Have I missed anything out?
If you’re interested in leveraging your data to discover meaningful insights or you want to save time and money on your reporting please find me at the event for a chat – it’s one of my favourite subjects.
Author: Scott Lawson, Trackpal