The Two Rules For Better Analytics

Do a search on Google on the term "better analytics" and you'll see 232 million results. In Amazon, there are more than a thousand books that reference better analytics. In this age of "big data" (however you define it), we all know that we are sitting on huge stockpiles of data, most of it gathering virtual dust in some analysis tool because we don't take action.Part of that is a fear of numbers. Let's face it: no one gravitated to Human Resources because they excelled at math in school. The word data implies a literacy in numbers far beyond calculating a tip or balancing a checkbook, a fluency most of us do not have.Part of that is a fear not knowing where to start. Data feels like the largest buffet you've ever seen. You can fill your plate with "what content did people look at most" or "what channels drove the most conversions?" but were those the most interesting and valuable items? Did you miss some amazing insight from traffic acquisition because you were filling up on content measurement and ad value?To those fears, I can say this: There are really only two rules for better analytics, and they don't involve any math what so ever.Rule Number One: Decide What You Want To Learn Before You Look At The NumbersBecause data is the world longest buffet, most agencies or analytics try to boil down an ocean of numbers into a summary report. It's designed less to tell you something meaningful or useful and more to show you that things are working. Like the dashboard of your car, its intention is to indicate at a glance that almost everything is going okay, but there might be something worth looking into more deeply. That red number on your report? It's closely related to the "check engine" light in your car: you're going to want to stop and ask some deeper questions.A dashboard can't give you insights because you aren't asking specific questions. If a coworker walked in and told you how many emails he responded to, how many phone calls he made and how many pages he wrote, is that a indicator that he is doing smart, productive and useful work? Of course not. It's a report of activities, not of analysis and insight.Numbers are just numbers. They mean nothing without context, and that context comes from asking it a question. For example, if I told you that traffic coming from Facebook has an application rate of 3.5%, is that good or bad? It's neither, because it's just a number. But if you ask, "Is Facebook performing as well as our other social channels?" then that number suddenly has value. It has so much value, it leads to even better questions like, "If Facebook is the worst performing social channel, what changes can we make in our campaign to bring it up to par? Or should we nix it altogether?"Asking smart questions is how you turn numbers into insight, which is why why it's our first rule.Rule Number Two: If You Aren't Going To Change, Don't Bother MeasuringI know plenty of people who claim to value their data. They collect it and study it, treating it like a valuable tool in their toolkit. Then, when it's time to decide, they choose to ignore the data. Usually, they say things like "I know all the numbers said "X," but I'd rather go with my gut."I know others who collect the same data and study it like there will be an exam on it, but then don't make changes. They either don't have the authority to make the change, or they are too scared to change.In either case, the time spent measuring and studying the numbers was completely wasted. In the first case, they were only looking for numbers to prove their point, not to inform it. In the second case, they flat out wasted their energy on something they couldn't do anything about. It would be like spending two or three hours of our work day worrying about whether the sun will rise tomorrow -- not the best use of your time.If you invest the time in your data, you should be reaping the rewards. Don't spend all that time exercising and dieting and then never go to the beach.Please note, that neither of these rules involved any math what so ever. They were simply guiding principles about how and when to dive deeply into the numbers. When the time comes for the really tough math, ask good questions and let your analytics team do the hard work of answering them.