We live in a time when self-checkout machines in supermarkets make cashiers redundant, we pay with our smartwatches and we talk to many smart devices around us that can tell us incredible things and even make us laugh. (“Alexa, tell me a joke,” I ask my Amazon Echo.) Yet, without the human touch, these awesome ways to connect to technology wouldn’t even exist and often wouldn’t work until someone came and pushed the right button. In a world where anything is possible, utilizing everything that is out there, without the right people, still seems impossible. It’s the typical story about man and machine. It’s the same for data. A piece of software can make it available to us, but if the digital analyst doesn’t know where to look, what metrics to compare or how to interpret what they’ve collected, it wouldn’t be worth a penny.
What does it Take to be a Good Digital Analyst?
Avinash Kaushik has the perfect formula for data analytics success: 90% of the investment goes to the right person with the appropriate skillset, the remaining 10% falls on tools. It seems exaggerated, but who am I to say he’s wrong. The man has breathed data for so long that comparing it with lovemaking is now natural for him. To be fair, talking about various data interpretations, features of data analytics tools and metrics in a sexy way makes me develop my own feelings about data analytics. 🙂
Another thing is that being a marketing analyst is a very passionate, extremely important and accountable job and many big decisions do and should derive from their work.
An analyst should have technical, marketing, business and analytical skills. They need an understanding of the business they are in and how to use the collected data to solve the specific problems that organizations face.
“On its own, data really is meaningless to people until they have the means to understand it and, as the digital analyst, you are in charge of providing those means.” (Chris Meares, What Skills Do You Need to Become a Good Digital Analyst?)
Without a story around the data, your audience won’t be able to understand it and, even less likely, solve any problems.
That leads to another point of caution. Creating a story shouldn’t be the analyst’s final goal, rather presenting the relevant data collected from appropriate tools, compared and correlated in a smart way, then embroidering those selected results to pinpoint the bad and suggest solutions for improvement. That being said, the analysts should never lean on their data to prove a hypothesis right or wrong. They should only use its “light” so they can find their way.
How can the Digital Analyst Influence the Quality of Data Analytics?
There are many examples to support the belief that the analyst is much more valuable than any data analytics tool. Here’s how them doing their job wrong can influence the quality of data analytics and thus unnecessarily dissipate resources and produce weaker insights:
1. Using two or more clickstream tools to solve different problems. This leads to spending too much time organizing and processing data, leaving the analyst with no time for actually analyzing different sets of numbers.
2. Avoiding A/B or multivariate testing in a real online environment through fear and waiting for case studies and success stories instead. It is much easier (and cheaper!) to try and fail than to wait for the ‘perfect’ formula that might even not work for you.
3. Relying on one tool only and avoiding multiplicity in your approach. There’s no one magical tool that does it all. Even a small organization has to employ several types of tools that provide different insights.
“The quest for a single tool/source to answer all your questions will ensure that your business will end up in a ditch and additionally ensure that your career (from the Analyst to the web CMO) will be short-lived.” (Avinash Kaushik, 10 Fundamental Web Analytics Truths: Embrace ‘Em & Win Big)
4. Asking too many questions in an obtrusive fashion, thereby showing disrespect for your users’ time. This usually results in not having enough answers instead of focusing on one problem at a time and gaining your users’ willingness to help you.
5. Collecting and archiving data that takes up too much space instead of reconfiguring log files according to the needs. This can lead to wading through meaningless data, increasing the time taken to analyze it and reducing the likelihood of successful results.
“Many data analysts subscribe to the notion that every possible data point should be collected so that different hypotheses can be tested. This is fine if you are conducting an experiment that will be difficult or costly to reproduce (such as sending people into space) but never the case when collecting web data.” (Eric Peterson, Web Analytics Demystified)
Conclusion
Let’s go back to the story about the man and the machine…
“The machine is powerful, but if the man does not know how to drive the machine there will be very little reward. It is not really one or the other, but it is actually a combination of the two that creates the winning formula.” (Silvio Ferrero, The Analyst is the Real Tool)
I have to agree with this one. There’s no doubt that the role the analyst plays in marketing and data analytics is invaluable. Being human, however, makes them prone to making mistakes. Even selecting an analyst is a human decision, so an employer looking to hire one should definitely try to avoid making an error with that.
P.S. In case you were wondering, this is the joke Alexa told me earlier:
Why are ghosts bad liars? Because you can see right through them! 🙂
1 comment