There are tons of ways that your organization’s data culture could turn toxic. This is a list of that won’t tell you the ways that your organization’s culture could go off the wagon, rather it is a list of the big things to watch out for. You can tell this in an interview, or you can tell this by seeing it around you. You will know that something needs to change. Not all of them will show up in every situation, and just because one of them exists doesn’t necessarily mean that your culture has turned toxic. But they can help you identify that the data culture has become toxic. So keep an eye out for these 5 signs of trouble.
1. The Learning Treadmill
Every year that I’ve been out of school, I feel like I am falling further and further behind the state of the art. Talking with many other data professionals, they feel the same. To a certain extent, you need to stay on top of major developments in the industry. But if you find yourself studying data, models, or other things related to your job in your spare time, all the time, you may be in a toxic culture.
This one is insidious because of hustle culture, and an ever changing industry. It is not unreasonable to feel like there are younger guns with better tools nipping at your heels trying to replace you. So you keep trying to learn more and more. And the more that you learn, the bigger your circle of competence becomes, and therefore the more you discover there is to learn.
The truth is that most of data science, is just cleaning data, moving it from this location to that, and then finally doing something moderately useful with the data. The more seasoned you become, the less worried you should be about the hotshot kid who knows deep learning, probabilistic programming, hadoop, and “theoretical econo-gymnastics”. The truth is that those kids probably need some serious guidance on how to actually provide value using those skills. A decent linear regression that can be deployed in hours is more valuable than a model that takes a month to develop.
If your management team expects perfection of knowledge in the latest and greatest techniques you may be in a toxic data environment. They should be focused on results, not on how you delivered those results.
2. The Top is Chasing Titles
Since I started my career, I’ve seen business intelligence specialist, data analytics specialist, and data scientist, all rise and fall in popularity. While at the same time data engineer and machine learning engineer is/seems to be on the rise. This is self explanatory, but if your management team keeps posting new job titles and hiring people with fancier and fancier job titles. They may just be trying to recruit talent using the buzzword of the week. But more likely than not, they don’t have a plan for the data function.
They probably, think that data is where value lies, but they don’t understand it, and they don’t know how to extract knowledge or value from it. So they keep hiring and hiring different job titles trying to find the person with the magic touch. They are hoping that someone somewhere can unlock the organization’s data and the value it contains. This is toxic, because, there is no plan, there is no strategy. It is just jumping from one thing to the next. If this is the case, take some time, develop a roadmap that will unlock value. You may be able to turn the ship around.
3. Unrealistic Expectations for your Model
Models are just abstractions of reality. They are stripped down and simplified to the point of absurdity.
If people are placing too much trust in your model. Or trying to use a model that you have developed for a different purpose. Or even worse, convincing others that your model can do something that it can’t. You are in a toxic data environment. When something fails, when something doesn’t work “as expected”, or if the model doesn’t deliver the insights that someone is looking for, you will be blamed.
The remedy is to be crystal clear that your model is wrong. All models are, but it should be useful for this very narrow sliver of things. If you can’t convince people of that, then you are definitely in a toxic environment and need to get out.
4. Unrealistic Expectations from You
This goes for any job. If you are always working, you will miss quality time with your family. If you are being crushed under the weight of work that should be for 3 or 4 data professionals. You are in a toxic environment.
I’ve fallen into this trap before. In a hustle culture world, it is very easy for your job to become everything and consume everything. Your job should exist to serve you and your goals. Not the other way around. If your job is controlling you, and it doesn’t serve your goals, then you are just a modern day wage slave. Get out! There are so many great companies out there that would love to have a talented individual join them. You can have it all, work-life balance, inner peace, and believe it or not, you can love your job. It does exist.
Life is too damned short not to love what you are probably spending most of your waking life doing.
5. Low Investment in Data
The final thing that you want to avoid is if data is a low priority via spend. If the company that you work for isn’t willing to invest in databases, and analytics tools to do your job, or refuse to hire enough data people. They probably, think that you are a wizard. Now there is a lot that you can do with open source technology that is really impressive. So maybe their opinion of you as a wizard is justified, but just remember, there may be less interest in using data. They may not know how to use that data.
In this scenario, you need to learn to justify the return on investment. If you can do that, and the investment is still low. Then you are probably in a toxic situation.