You'd think there would be a simple, straightforward answer to this question. Even allowing for differences between companies and working styles across different industries, surely there is a set of rules and benchmarks we can measure managers against and see what makes the best ones great?

It's not exactly rocket science after all.

Pop into a few meetings, take the data requests in, dish the work out, arm around the shoulder if needed, kick up the metaphorical arse otherwise, butter up the execs and senior managers, claim a large slice of the reflected glory, rinse and repeat.

In many places that's probably not too far from the truth and I could cut this piece off right now and call it a day.

But it's not all there is to the story. Especially not if we aspire to be a better kind of manager. What we really want to know is what separates the good managers from the not so good ones, and what makes a good manager into a great one?

More so, if you want to take this career route yourself, what do you need to do to be one of the really good ones?

What does great look like?

Caveats first: opinions differ depending on experience, type of organisation, culture and industry. I read a lot of forums, message boards, and social media threads and the advice ranges from one extreme to the other.

I was a senior analytics manager for five years so I've been in the hotseat. I've also been back as an IC working with a range of different managers for the past seven years. Different styles, different approaches.

But my experience on both sides of this fence has led me to see the following as the main high impact drivers a data manager can bring to the table, regardless of where they are working.

Servant Leadership

I'm going to kick it off with this one as it underpins my personal philosophy on all of this these days: management works for the ICs, not the other way around.

It all comes back to the concept of Servant Leadership

A servant leader shares power, puts the needs of the employees first and helps people develop and perform as highly as possible. Instead of the people working to serve the leader, the leader exists to serve the people.

This applies to all levels of management and leadership in the organisation but, from my own vantage point, it's especially relevant to the relationship between a line manager and their Individual Contributors (ICs).

Without getting into biblical and historical critiques of the works of Jesus Christ, Herman Hesse and Martin Luther King Jr. (damn this article got heavy really quickly...), how can we actually go about being a servant leader in our work as a data science manager?

Let's dig in.

Be A Pushback Monster

If you have one main role in your job as a data manager, it's to be the air defence system for your team, no matter what is being thrown at them. You have to get in there, intercept the ad-hoc unplanned requests and questions, and cut them off before they get near your people.

Knowing how and when to push back like mad on those data-hungry stakeholders who would suck up every last second of your crew's time with question after question after question is a golden skill.

You might be conflict-averse, many of us are. Not everyone you come up against will be though. Some people love to get into a good old to-and-fro and throw their metaphorical weight around when they want ("need") something done in an urgent fashion.

Learning how to deal with all of these situations will be a make-or-break position for your future as a successful manager. No pressure. Fail to protect your team and let other people walk all over you to their detriment, and you'll never recover. Remember always that you work for the team. Be their shield.

Removing Roadblocks

I've heard it said that good managers react to blockers quickly and decisively, while great managers prevent them before they occur. And I'd have to agree.

These first two points are going to make it sound like being a manager is a constant series of ever-increasing battles and wars with people working within your own company. With friends like these, who needs enemies?

It's never quite as bad as that (hopefully).

But you do have to become adept at conflict resolution, stakeholder management, and back-and-forward communication. There is no way around it.

Roadblocks might come in the guise of a dependency on an engineering team who do not share your level of priority on getting the work onto their roadmap to unblock you.

It might be from your own senior management deprioritizing your additional headcount request for an important big project. Or the loss of a team member for an extended period of time for health or personal reasons.

We can't always foresee these bumps in the road before they happen but the great managers have either contingencies in place or, at the very least, can rustle up a Plan B to get the team over and around them.

Seeing The Big Picture

Dishing out the work as it comes in (after filtering out the ridiculous and unnecessary elements of course) is always going to be a big part of the job. We don't ever advocate for becoming a Data Vending Machine where users punch in their request and your team just spits out the requested results. But, depending on your remit, that will sometimes be inevitable.

What will really set you apart as a great data manager will be the ability to step back from the day-to-day business-as-usual hubbub, and see things from a higher level across the whole piece.

Understanding the business will matter more and more as AI takes on more of the technical aspects of a data team's job. This will apply just as much to you as a manager as it will to your ICs. What an LLM can't do is build the personal relationships across different functions in the business, mould all of the things you learn from those subject matter experts together, then translate that into a vision for how your data science team can better support the business in future.

Not only does this give you a better position to suggest new ideas and initiatives, it also helps you understand how your team's work relates to other teams, and then make the important calls on prioritisation before conflict happens.

A visionary AND a pre-emptive conflict resolver? Get that Nobel Peace Prize over here now.

Promote your team's work

There is nothing more craven and stomach-churning than the manager who takes all of the credit and glory for the hard work of their own team, and passes none of that on to the people who made it happen. Even worse if they use it as a lever to forward their own career and chase promotions.

Down with that sort of thing.

You should always be striving to promote the impact of the work your team is doing. Reward them. Give them shout outs where you can. Shine a great big bloody spotlight on the work they are doing and do your absolute damnedest to make sure the bigwigs way up in those golden offices at the top see what they did AND know who was responsible.

Don't get me wrong, it's fine to stand at the side and make sure people know this couldn't happen without you. We all play our part and, if you've done the other parts well, the well-oiled team machine should be excelling because you helped them do it.

But don't be a glory hog. No-one wants to see the coach lift the championship trophy at the end of the season. Leave that to the players. They'll respect you more for it and you'll ultimately have more respect for yourself as well.

Developing your team

This is a tough one. I wrote an article about how becoming a manager isn't a promotion, it's a career change and it's still a very valid point to make for anyone considering the switch. Your own skills growth will now be going off in a very different direction to where ti did before.

Coaching your own ICs is now your main focus. Where can they improve? In their tech skills, "soft" skills (I really hate that description, more in a later article), personal development, stakeholder management, the list is endless.

And if you get really successful and make them into a proper superstar, maybe, just maybe, they'll ditch you for a better offer elsewhere at another company. Probably make a hell of a lot more money too the ungrateful so-and-so's...

And so they should! This would be a successful scenario for you. You'd have played your part in helping make someone's life better. You might have also made your own a little more difficult for a spell but this is the job. You do want to be a Servant Leader, don't you? Of course you do.

We're team players. We work for the team. We develop the team, one by one, and as a group. That is the job.

Do (A Little) Yourself

When I moved into management, I wasn't quite ready to give up the old IC world altogether. I still liked to dig into a dataset and poke around in there. Analyse this. Analyse that. Maybe put a data product together. Build a pipeline or two.

So I did.

There's no rule that says you can't keep on a little IC work if you want to and it doesn't interfere with being the roadblock destroyer and Air Defence squadron leader for the team.

Many ICs actually prefer if their manager does actually retain some of their technical chops and is willing to roll up their sleeves and get into the actual dirty work alongside them.

It could be pairing with junior analysts or data scientists, doing code reviews, taking on small tickets or requests - anything to keep yourself up to speed and your knowledge relevant to still be able to talk shop to your team.

Some of the best managers I have known were able to be hands-on technical but also knew how to speak the language of the business as well. I don't believe you can properly do one without the other as a data science manager. You straddle the two worlds.

You don't have to be the best DS on the crew, it would be a failure of hiring and development over time if you were, but you can't let your knowledge slip too much or let it go completely.

It could be stats knowledge, or coding, or data engineering, or analysis skills or...hell anything that fits under that massive big umbrella we call "data" these days.

Just enough education to perform, as Stereophonics might have put it, will keep you right.

Sounds simple, doesn't it?

Obviously it's not or every single manager who took the job would be a great manager. And they really aren't. Oh my word, so many are not cut out for it at all.

And that's ok if they catch that on quickly enough and hand over the whistle and clipboard to the next coach who might just be a better fit. It unravels quickly if they don't though. We'll cover the main actions and behaviours you definitely don't want to experience in a data manager in another article.

If you're already in a data science or analytics manager job, or think you might be interested in taking that career change decision, I hope these few suggestions for areas to focus on are useful and give some food for thought.

As I said up above, these are some of my personal beliefs in what that job should mean, but I know that's far from a universal approach. The good thing is you can always take what you've learned from across the wide range of your life and apply it here within the context of your data team.

Maybe you had a great teacher at school or a sports coach that stood out in the way they handled roadblocks, setbacks, and challenges.

Maybe it was a manager you've served with in another job and thrived under. Or maybe it's a really horrible one that you've decided you'll do the exact opposite of when it comes your turn. It all counts.

Good luck however you approach it but always remember to put the people first, above all. You'll rarely go wrong with that as your real North Star.

(Photo by Clark Tibbs on Unsplash)