TTE Strategy named HIDDEN CHAMPION 2022 / 23


Wiebke Apitzsch

You want to use data for your business model? consider these 3 tips


Over the years, I have supported many different companies in various industries with their data strategy development. Due to the uniqueness of each organization, companies are facing different challenges when trying to use data to optimize their business model. But there are three stumbling blocks in particular that every single one of my clients has had to deal with sooner or later.

To avoid them, I suggest that everyone in charge of an analytics project should consider these three maxims.

1. Analyze for good reasons and not for analytics sake.

First, you need to identify the questions that really matter. It is only then that you can think about the way to get your answers. Data analysis is one of many tools you can leverage. While it is powerful in many situations, it is not the answer to all questions. I love analyzing things, and I could sit and find you patterns for days, simply because it is fun. But running analysis isn’t always the best way to solve a problem or to find a new direction. Therefore, an analysis needs to follow a clear data strategy - it requires a definition of done.

2. Insist on proper data cleansing.

For me, the fun part in analytics is the dashboarding, the deep dives, applying fancy methods or running different methods to see what works best. But I’ve heard and said that sentence too many times:

Garbage in – garbage out!

There is no good analysis based on incomplete, inconsistent or biased data. Data cleansing is hard work, it is no fun, requires a lot of communication and no one praises you for doing it. But, without it, all fancy reports are pointless. I can say that in my professional life, I have never received one complete, cleansed, usable dataset to work with. If your analyst does not come up with questions about your data after having a look at it, you’d better be careful to use their results.

3. Hello Culture Clash:

When Sheldon Cooper meets Oprah.

If you’ve ever watched a movie at an American college, you were probably not surprised to see that the tecchies dressed differently, had different parties and talked in a different way to those who studied arts or business administration, or law, or any other field. Now, there are always exceptions, but I worked in IT and Business and I think it is fair to say that each group has, generally speaking, their own culture.

Now, no one is really surprised about the students in the movie, but in business you seem to expect that you just put two people with completely different lives, communication styles and experiences in a room, ask them to be professional and then assume they can figure out how to communicate on their own.

Here is what happens most of the time:

The business guys know how to talk, make assumptions, and get things going. They learned how to show confidence, even if they are not 100% sure. They have to do that - most of their job is making good decisions based on incomplete data, just because there is no certainty in business. You decide for one direction, and no matter how it turns out, nobody can really tell if the other option was better or not.

 For analysts, things are very different. People do win the lottery; irrespective of the insignificant likelihood. But, in contrast to business, there is such a thing as correct and incorrect in analytics. Wrong factor, link missing, uncleansed data? Someone will crosscheck your numbers, and if you’ve made a mistake, they will find it. It is 1 or 0, correct or wrong. You have to be cautious. You always say: I guess it works fine now, but please feel free to crosscheck, I am not 100% sure yet. (And I never will be, people do win the lottery, you know)

I mean, you only spent two weeks calculating the numbers back and forth, till now, all results were accurate, but maybe there is one combination of settings that could still break it? Who knows? You’d better write a disclaimer.

Now, as often in life, you have a very confident person with limited knowledge and a very knowledgeable person with little confidence.

What do I do in business to find out if the one I need to trust is trustworthy? I google their topic really quickly, challenge him a little and see how certain the other one puts me in the corner. If he can, it is cool, else he maybe knows less than me, which is bad.

As an analyst, I have a different experience. Here is what you learn, either the hard way or by listening to others:

“Don’t you ever mention the name of a new and fancy method in front of a business person.”

The moment these guys hear a buzz word, they catch it, run away and then never let go again. Even worse. They go and tell their boss “We will leverage natural language processing for this” and everyone goes like “Oh, cool project, he is innovative, impressive!”

I can say that because I work in both worlds and I ran away with buzzwords before, and yes, I would do it again. These ideas sound so cool, who would not want to do a cutting-edge project?

And the analysts sit like “No. I said we will NOT use it. It is 15 emails only. You could just READ them, couldn’t you?”

So, I think you get the idea.

On the other hand, here’s a pro tip for analysts: Executives do not always react positively if you try to crosscheck if you can exclude Cyrillic symbols and Arabic writing for an analysis for a German company. I was once surprised by that, and, yes, just because you’re paranoid doesn’t mean they’re not after you. So, I am still asking, but they never appreciate my diligence.

In the end, what needs to take place is that both parties get to know each other, learn to value each other’s strength and build up trust so a real conversation starts.

Sound like a funny story? You won’t belive how many times I have seen projects fail, just because of a proper culture clash between the tech and business world. So make sure to bring someone into the project who can translate between them, understands the demands from both sides and is aware of the different working styles.

To sum it up: Develop your strategy, whether it's data-driven or not. Be pragmatic about using analytics and see if there are easier ways to get where you want to go. If you decide for a data-driven approach, look at what exactly you want to know, insist on good data cleansing and then make sure the team has a good start, ideally with the help of someone who understands both.



The Team


// Google Tag Manager (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start': new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0], j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src= ''+i+dl;f.parentNode.insertBefore(j,f); })(window,document,'script','dataLayer','GTM-5B3DCSJ'); //Hubspot var script = document.createElement('script'); script.setAttribute('id','hs-script-loader'); script.setAttribute('src','//'); document.head.appendChild(script); Privacy Imprint