9

I am a data scientist in a big company (10K employees).

I have to say I appreciate my job (for now) and leaving is not my preferred solution, even though it's in scope.

So, in the last 2 years, I built from scratch a brand new internal Data Product for my company. A few Data Science colleagues helped here and there, but :

  • I am the only one completely mastering the pipeline behind from A to Z (and it's not just an algorithm on the shelf)

  • This deep knowledge allow me to design very relevant and innovative data studies for my company, and I've done it 100+ times now

  • I am now the most experienced in the company on the scope adressed by the tool, almost becoming as expert as my expert colleagues the tool is supposed to assist.

My company is pleased with the current situation because the tool proves useful, and they have been really nice to me which is a great advantage of being a useful bottleneck, but I slowly realize that being a bottleneck comes with drawbacks too.

In particular :

  • They won't fight hard to give me career progress, as I am really useful where I am...

  • The tool still has much untapped potential, it could be used by 10 times more people within the company, could be more automated, I have untapped awesome R&D ideas, etc etc... but I just lack time and resources to do all this while spreading it in such a large organization. I was alone when the tool went from 0 to 1, and it was fine, but it becomes frustrating to see no one is able to help me make the tool go from 1 to 10.

Do you have any experience to share where someone in a IT position was a bottleneck and managed to get past the time trap ("have no time to improve current situation") ?

Edit : N+1 has been in the exact same position for the past 10 years, sees the results of the tool, finds it great, but he has no team vision, does not like his management role (prefers being hands-on things than manage people stuff), unfortunately has very few technical skills, so all in all he feels clueless when it comes to help me. N+2 is almost the same, and N+3 is the CEO.

2
  • Where's the "pain point" motivating the company to optimize the bottleneck and expand your group? Where is it for N+1? If there isn't one, perhaps take the line of "I would like to develop career and have people reporting to me, considering other options if stuck"
    – Pete W
    Commented Dec 14, 2021 at 13:29
  • How much training would be required for someone else to take over the position of "Data Product Operator"? Obviously, in order to step out of the role, you're going to need to get someone else to step into it, but how hard is that going to be? How doable would it be for you to adjust the product so that it was easier to learn/use?
    – Ben Barden
    Commented Dec 14, 2021 at 19:58

4 Answers 4

14

From the way you frame your question, I'm not sure if you know what you want yourself. Clear that up first. What is it what you want? Work on the tool? Do something else? Transition to team/department leader and lead a team of Devs working on your tool?

After that, talk with your direct manager about it. If he is not supportive, ask why and gear up. Come back to him with arguments. If the direct manager can't decide, go up the chain of command. But go together, don't just go over his head.

If everyone just wants to preserve the status quo, you also have an answer. Then you can decide if you want to pursue opportunities outside your organization.

2
  • 3
    and if they want to preserve the status quo, you should have a fair amount of leverage for raises because you are the bottleneck. Just because you're not becoming a manager/etc doesn't mean you're not providing the company good value. You should be getting compensated for being a critical person. Commented Dec 14, 2021 at 14:04
  • You are right. I spent last week answering the question "what do I want". Now I have objectives for my tool, and consequently objectives for myself and for my company. Let's see if we all can make it happen Commented Dec 20, 2021 at 9:05
1

IMO learn to trust and delegate. And take it slow.

If you want your business to grow, you have to take on more clients. If you take on more clients, you need to hire help to deal with the extra business.

In your case, if you want your project to grow, learn to trust others and teach them how to help you grow it.

Organise pizza parties (gonna be harder if remote work is involved) with colleagues you think have skills and are useful to implement the improvements you want. Try to sell them on joining your project as a temporary thing. If they are interested, involve your managers to talk to their managers and have them budget some time for your project. It doesn't have to be a lot at first, just a few hours every week just to see if it works out for you.

Ask your managers to help you in your search for talent if you don't have anyone in the company that fits your needs. Present them the opportunities you've seen and give them a cost/benefit analysis for throwing more brains into your project. If they are technically inept, and let's face it, most people in management position will be when it comes to DS, then get involved in the selection process (help with the tech interviews, prepare presentations for colleagues or external candidates, go to Universities or wherever potential talent might lie untapped and try to reel them in).

Understand that you don't have to do management work if you don't like it, you already have 3 managers above you that should be skilled enough for that. Learn to delegate all the boring managerial tasks to them, but maintain agency for the code, features and talent that you want in your project.

If you get enough help to get your project from 1 to 2, try "selling" it across the company. Set up small meetings with managers from other teams that you think might benefit from your project and give them a non-technical presentation of how it will help them get more work done. If anyone bites, work with their teams to integrate your product in their workflow. If it's a success, ask those managers to quantify how much money/time your product saved them.

TL;DR version: Don't try to do everything yourself, use those resources available to you to gain visibility and support, then build from there.

0

If you have faith in your ideas and the ability to see them to fruition you can always work for yourself. Untapped niches are hard to come by but lucrative if you can get them in service. You cannot force a company to support you but you can support yourself.

It's not the easy or safe way though.

-2

Case of stale management

Your company has a bad case of stale (stagnant) management. Imagine a manager coming to work, and knowing everything is working more or less fine and as usual in his department, proceeds to watch football. Or chat with his mistress. Or reads Plato. Or dabbles in occult magic :) In any case, his mind is not on his job, and he fills satisfied and secure in his position. He does not need more money, or better office and businesses car, he does not feel threatened by his superiors or his subordinates.

And yes, your management sees you like a screw in their machinery . Useful, yes, expensive, probably. But still just a screw. They occasionally oil that screw (with more money), but would like to keep it at the same position until retirement.

What can you do ? Well, if you don't want to stagnate there are only two options :

  1. Try to ruffle some feathers. Go directly to your CEO (N+3 level) and try to interest him with your ideas. He might get upset, your N+1 and N+2 could get upset, but at least you tried ... There is a small chance CEO would see some potential and allocate you more resources for your job.

  2. Prepare to abandon ship. In most cases of companies with stale management, it is not worth the effort to try anything. Simply start looking for another opportunity, some other company or even your own startup (alone or with friends). When the time comes, just give your notice, say thank you and smile, do not fall for emotional blackmail and leave gracefully.

Not the answer you're looking for? Browse other questions tagged .