If you are reading this article, then you probably care about data-driven thinking and like to make decisions supported by data.
Not everybody at your organization is like you!
I’ve worked with a lot of online businesses as a consultant (or as a full-time data analyst) in the last few years and I found one tricky question coming up from time to time at many, many companies:
How can you evangelize data-driven solutions in an organization where there is no established culture of using data?
Sounds like a nonsensical question, right?
The goal of a data science project is to support the work of everyone at the company. It actually helps people! Learning about your customers and reacting to their needs is the one most obvious thing every online business should do all the time.
So how come there are always people who are opposed to – and sometimes even protest against – the data scientist’s findings? How is it possible that when you present your results, there will be always someone who “knows better” in her guts?
There could be many reasons for data resistance. First of all: personal interest.
You say: “Data shows this, so let’s change that.”
They say: “I don’t get it, but let’s do it like we did before. It worked, so it will work in the future, too.”
If you are a great data scientist, you will bring some game-changing ideas to the table. And change is scary. Especially for those stakeholders who are really confident about the things they are already doing. When you say there is room for improvement, they take it as a personal attack against them and their work.
Don’t worry! At the end of this article I am going to share my best practices on how to deal with this kind of issue when you start transforming your company into a data-driven organization.
The other reasons for data resistance
It’s also worth mentioning that the data resistance within an online business is not necessarily one specific person’s fault. It can come from the lack of data-driven culture and lack of communication, too.
I’ll give you a very specific example. In 2015 I worked with an e-commerce company who primarily sold computer accessories. Their business was very simple: they got products from the manufacturer and sold them in their online shop. The producer sent photos (usually low quality photos) and descriptions (extremely boring descriptions) for the products.
The e-commerce company wanted to have nicer things on their website. Their hypothesis was that higher quality would bring better conversion. Not a bad idea, right? Anyway, they decided to take it to the next level and they just kicked off their “quality uplift project.”
They wanted to replace all the images with high quality and unique images and all that boring copy with easy to read and more exciting descriptions. So they hired two freelancing copywriters to write content and two freelancing photographers to shoot professional photos about the products.
After three months they hired me to find out how the project was going.
Unfortunately all my research and AB-tests showed one thing very clearly:
Though the new better and nicer photos were working well, none of their customers read the new more creative product descriptions… It turned out that the customers didn’t want inspiring and well-written copy.
They cared about two things: price and specifics. (Interesting, right? The original boring description converted better!) The new wording didn’t make a difference in SEO (search engine optimization), either.
Don’t get me wrong – the copywriters really gave their best. They worked hard and produced exceptionally great stuff. But their work didn’t support the quality uplift project’s original goals (better conversion and more sales). However, it wasn’t their fault! The bad decision happened on a strategic level 3 months before.
Hmm… So now what? (Hint: they didn’t get fired. :-))
My point is: as a data scientist, you can’t go to a company and change everything at once.You can’t say to people that their job is useless or the things they were doing were all wrong. (After all: conversion rate and money is not everything, right?)
If you do, you might end up with one of these scenarios, and you won’t like them:
- Talent will leave the company and the company loses good people. Not because they didn’t do their job right, but because they didn’t have the data at hand when they needed it.
- The data scientist (you) will leave the company because everybody will hate you. Either you get fired or you quit yourself.
- Everybody stays at the company, but data resistance will increase because of personal issues.
None of these sounds too good, right?
How can you break down data resistance? How can you grow the data-driven culture?
I promised to share my best practices, so here they are. These will help you to make sure that you can do your job as a data scientist efficiently and effectively… and at the same time keep everyone happy at the organization.
- If you have a small online business (e.g. startup, e-commerce), hire a data scientist in the early phases. (She should be around your 20th team member – or you can hire someone as a half-time employee when you have a 10-people-size company). Having a data scientist on board from the first years will help a lot to shape everyone else’s data-driven mindset and – in general – the company’s data-driven culture, too.
- On the other hand, if you are a data scientist and you are about to join an online business as their first data professional, try to join a smaller one where you can still shape the data-driven culture. If you like challenges, you can work with bigger organizations as well (like I did several times). More friction, but more impact eventually, if you do things right.
- Start with a small side project! On less significant projects there is lower data resistance — well, a lower chance that you’ll step on someone’s toes. Once you are done and your “demo” data project was successful, you can use this project as an internal reference in the future.
- Find multiple channels for communication! Slack, email reports, presentations, workshops, etc. You have to build up the “internal marketing” of data-driven thinking within the company. As a good evangelist, try to attend every important meeting and ask before every decision: Why? Why do you think this is a good project? Do you have any research supporting that idea? Have you tested it before?
- Educate! Set up 1-hour workshops, where you show people what AB-testing is, what user testing is, what heat mapping is, what predictive analytics is, etc… People love to learn. Especially when it’s free and during work time. 😉
- Repeat things! I had several cases in which I presented the same presentation about one data project almost 20 times in 2 weeks. It was worth it in the end: it’s better to spend 20 extra hours presenting the same thing over and over again than throwing away a project that took 200+ working hours because nobody knew about it.
- Do one-on-ones! A lot of one-on-ones. Sit down with the people who don’t understand or don’t agree on things. Explain everything in detail! If they don’t have time to sit down with you, don’t get mad, just make it clear to them that they have the opportunity to have a one-on-one with you anytime.
+1: Be patient! Building a data-driven organization (or transforming a company into one) is not an easy job. But be persistent, look at it as your personal mission and sooner or later your colleagues will turn. I have seen this many times and I can tell that once someone gets a taste of working with data, she will never look back…
As you can see, breaking down data resistance and evangelizing data-driven thinking is not an easy thing. Neither is it friction-free or fast. It could take weeks or months. But the more successful data projects your colleagues see, and the more great results you deliver, the lower the data resistance will be — and the happier, more effective and more successful your data-driven organization will be.
I guarantee that.
- If you want to learn more about how to become a data scientist, take my 50-minute video course: How to Become a Data Scientist. (It’s free!)
- If you want to learn everything that you have to know about A/B testing (business elements, science elements, best practices, common mistakes, etc.) and become a real pro in building winning experiments, take my new online A/B testing course called A/B test like a Data Scientist!