Education. Time. Budget. Resources. Foresight. Successful measurement and analysis is prone to numerous roadblocks. (What did I miss?)
The big upside to the four problems I named: in an ideal situation these problems can be solved:
Education. In our ideal scenario, your team understands the value of measurement and how to do it correctly. Your team knows to bring in the analysts at the beginning, at the end, and throughout the process to ensure it’s done right. Everyone is sure to think about measurement and sets SMART objectives which you use to choose metrics. Everyone agrees on the definition of success and what data will be used to show it.
Time. You have all the time you need to do the job right. You’re not under an impending deadline or you’ve ensured that your trusty analyst has all the time they need to crunch the numbers and provide you with what you need.
Budget. For some reason, money is not an issue for you. Your team has planned for measurement in its budget or you have an endless supply of cash (or both). You can afford that fancy tool you want to use, and you can afford to pay for all the analyst time you will need.
Resources. You have a well-trained analyst and a solid team. You can easily articulate how you measure success and how it needs to be done. You have the right people to make it work (this goes along with education).
Foresight. Measurement was on your mind from the start. You have set up benchmarks and pre-campaign metrics so you have something to compare to when all is said and done. You have set up the right tools to start capturing data as soon as it’s available. You don’t find yourself at the end of your campaign scrambling to pull together the right pieces, you’ve already planned and compiled them into one place.
While all of the above problems are frustrating and taxing and often enough to cause you to stop trying, they can be solved. And I know, it’s probably unlikely that you’ll be lucky enough to find yourself in the ideal situation. But it is possible.
There is one even more common problem that is not (currently) solvable: attribution. We have to make a lot of assumptions in measurement to be able to quantify value or success with any kind of regularity.
We assume if I have 2,000 followers, each time I tweet I reach 2,000 people. We assume if I click on a banner ad, I’m interested in the product or company to some extent. We assume if I click that same banner ad and end up buying a product, my decision was fueled by said ad.
But what if I tweet at 3 a.m.? How likely is it that more than 2 people will see my tweet? What if I clicked on that ad because my mouse slipped when I was trying to click somewhere else? What if I bought a product because my friend recommended it to me two weeks before I happened to see that ad?
And therein lies the rub. There is (currently) no available method to track attribution reliably, consistently and accurately.
Even if you exist solely as an e-commerce website and only use banner ads to advertise, there is no way you can prove that I bought your product because I saw your ad. We all know how powerful word of mouth is, we can’t discount it and we often can’t measure it.
But don’t be discouraged! Don’t give up measuring! Focus on solvable problems, and do what you can to skirt your attribution problem (more on that in another post). Understanding your limitations and what you can (and can’t) measure will only help you build better and more accurate measurement programs.