What's your opinion?

Why opinions matter and how to find yours

I was surfing LinkedIn today and saw a post referencing Klarna’s CEO. The content of the post wasn’t important; however, it finished with a line about the importance of opinionated tools. It stuck with me.

My career in data has been punctuated by tools with strong opinions; Tableau had the strongest of them all. Just watch this short excerpt from Christian Chabot’s keynote in 2014 at the Tableau Customer Conference.

I don’t know about you, but this gives me goosebumps. Chabot’s Tableau had a clear, strong message—creativity is at the heart of good data analysis. It gave direction to the tool, their platform, and everything they did, and as users, we bought into it. Alteryx, too, had a strong mandate and opinion—perhaps not as strong as Tableau’s (and you may argue heavily influenced by Tableau’s message)—but they set out to make data processes simple and easy to automate while empowering business users with data.

Both these tools have inspired a slew of tools in their wake. However, I’ve never really latched onto tools in the same way. Perhaps, recently, dbt’s focus and opinion around driving software development methodologies to the heart of data processes is the opinion that has most recently made me sit up and listen.

Compare that to the opinions of tools like SQL or Python. Or even tools in the Microsoft stack—when did any of those tools truly have an opinion? I’m not criticising them here; I just mean it’s harder to hang your hat on what exactly these tools are about.

Why Opinions Matter

Why is this important? I think, at the core, the simpler the opinion, the easier it is to latch onto and be guided by its principles.

Tools with opinions guide you straight to best practices—you don’t need to reinvent anything. Opinions are your guiding strength here. They make things quicker and easier to adopt; the tool itself and the philosophy guide you to the best way of working.

At the same time, we experience less decision fatigue compared to using more flexible platforms. We can quickly make decisions and know the right approach simply because there are fewer options in the tool. For example, in the early days, the dashboard owner in Tableau drove the decision around who could access the workbook in the platform. This philosophy was at the heart of Tableau’s ethos around ownership and creativity, removing the need to decide the appropriate governance model. [This has since been expanded to give more control to the business and introduce tighter governance.]

Of course, there are trade-offs—less decision-making means less control. You have to work with the tool; if you work against it, you’re fighting the opinion, and you’ll get frustrated. If you have niche needs, this can be extremely difficult.

Why Does This Matter to You?

Data tools seem to have lost their opinion lately. In a crowded market, they try to be generalist tools that please everyone—perhaps opinions are the preserve of up-and-coming challenger tools rather than incumbents.

However, that doesn’t mean you should lose your opinion. As a data leader, having an opinion and wearing it on your sleeve helps guide your team and colleagues. The simpler it is to state your opinion, the easier it is to understand, and the easier it will be to follow.

Even at an analyst level, I talk to my mentees and colleagues about developing opinions early in their careers and speaking about them—both blogging and in person.

Does your opinion matter? To a degree, I’m not sure it does. As we noted above, what matters most is having a guiding principle that helps your team understand best practices and make decisions quickly.

Here are a few examples of strong opinions that I’ve encountered in my career:

• “Nothing leaves the team without being checked—no excuse, ever.” This was my first manager Simon’s mantra. It was repeated all the time, and we lived by it. As a result, when we were being rushed by the sales manager to get some analysis out to a client as soon as possible, we knew what to do. We slowed down, we got a colleague to check the data, and we made sure it was right. I happen to think Simon’s principle was the right one, but it doesn’t really matter—what matters is that the team pulled together and approached things the same way every time. Decision fatigue around this issue was zero.

• “All our enablement/support queries go through the Teams channel.” People got fed up with us saying, “I’ll answer if you can get it on the Teams channel,” but it worked. Engagement in the channel went up, and more people saw questions and answers from their colleagues.

• “We only work on projects that will generate £1 million of revenue or above.” Rightly or wrongly, this drove the team to find high-value projects to save the business money. They knew to say no to low-value projects. [Again, I’m not commenting on the opinion—just the fact that it drove an outcome and reduced decision fatigue.]

What’s Your Opinion?

So, it’s important to have an opinion—it should be at the heart of your data strategy, and it should be simple enough to articulate easily and for people across the business to understand it.

Are you driving speed over perfection—aiming to help your customers fast and gain a reputation for iterating and providing answers quickly? Or are you building a robust, documented platform that serves as a solid base of data at the expense of short-term wins? Do your team know? Where flexibility is needed, do they know when your opinion will bend under stress?

Are you driving a product mentality with data? What does that mean to you?

Are you driving automation over customisation? Where is the line?

What’s more important to you—data or value?

Do you want your business users to become data users or data consumers? How are you enabling your business users?

I hope you know your opinion on these items. But if I asked your team and board, would they know? Wearing your opinion on your sleeve, backed by reasoning, is an integral part of [data] leadership. And if you’re not there yet, are you growing in your role in data?

So, tell me your [data] opinions—and would your team agree?