If you search the web for ‘first startup data hire’, you may come across some strongly-worded advice claiming that this person must not be a data scientist, or that they must be a data engineer / analyst. In my view, being so prescriptive about titles risks missing out on great candidates. The reality is that titles in the data world are messy and fluid – it’s best to start by getting clear on what the data person is going to do, and proceed from there.
Beyond titles, this post summarises my perspective on questions that arise around the first data hire, and presents some pointers to help you hire successfully.
Assumptions and Timing
Key assumption: If your startup needs a data hire, you’re probably at a stage where you’re starting to be limited by visibility into your data. You are generating revenue, but your data is all over the place (spreadsheets, dashboards of various tools, unstructured logs, etc.). There are important questions about your business that you can’t answer because you’re either not collecting the data, or because it’s too hard to gather it into a coherent story. No one on your technical team has informed opinions on tools to pick out of the dozens of options for warehousing / ingestion / transformation / analytics / orchestration.
If this is the case, then data and machine learning isn’t core to your product. You need someone to set up your data pipelines and analytics. These will primarily serve internal-facing use cases, like driving marketing decisions. However, the first question to ask is: Do you really need to hire someone for a permanent full-time position?
Personally, I’m biased in favour of not hiring (yet): You can get started on your data journey with a contractor or a part-time (aka fractional) person. This should give you a better understanding of your data needs, and get you to a better place in terms of data infrastructure and dashboards. This person may also want to become a full-timer down the track, or help you with hiring other data people.
Remember that – by definition – premature hiring unnecessarily shortens your runway. Hiring and onboarding a full-timer would usually take longer than bringing on an experienced contractor. And if you need to let them go, it may adversely affect team morale. This doesn’t apply to contractors, who are expected to leave when their contract is over.
That said, there is value in retaining a long-term owner of your data and analytics. Every business, dataset, and data stack have their quirks, so the familiarity that comes with long-term ownership is a definite point in favour of hiring for a permanent role. That said, you should still be open to part-time if you don’t have full-time needs yet.
Titles and Skills
If you do decide to hire for a permanent role, there are three other articles worth reading:
- Andrew Bartholomew covers assumptions (similar to the above), responsibilities, skills, management, and the thorny question of titles. He says that the person’s title is “the least important question […] you’re hiring a Senior Analytics Engineer or a Senior Data Analyst, but in practice this person might prefer a Senior Data Scientist title, or Analytics Lead, or something else.” I agree with this and pretty much everything else in Andrew’s article, though it is important to align on the expectations implied by titles (more on this below).
- Colleen Tartow advocates for hiring a senior data engineer. While Colleen’s advice is sensible, I’d be careful with following it blindly due to the messiness of titles and experiences. For example, you probably don’t want a data engineer who’s only worked with big companies, as there’s a risk that they’d over-engineer your data stack (initially, you’re aiming for a minimum viable data stack). Also, if they’ve only ever worn the data engineer hat, they may find it hard to uncover and communicate the insights you’re after.
- Sebastian Hewing goes deep into the question of timing the hire as a function of product-market fit. I agree with most points, but disagree with this phrasing: “The last person you want is a Data Scientist. […] What you need, in my opinion, is a Head of Data & Analytics.” I believe that someone who has full-stack data science experience may make a great Head of Data & Analytics – it all comes down to skills and experiences rather than past titles, which can only ever tell a part of the story. That said, Sebastian does list a bunch of other data titles that the startup shouldn’t hire, so we probably agree on the essence of the role and the person. I especially like Sebastian’s emphasis on seeking a hands-on data person who can turn data into insights AND insights into action.
As you can see, the three articles disagree on the question of titles, with Andrew’s being the most pragmatic. If you want to get even more confused, ask ChatGPT to summarise the collective wisdom of the internet: When I asked it “what should a startup’s first data hire be?”, ChatGPT suggested seven(!) roles with an “it depends” reason for each one. Personally, I’d go for a senior data generalist with an engineering background, who is also attentive to the business side. It’s highly doubtful you’d find someone who goes by this title, so you’ll need to figure out how to find and attract them. This is hard if you’re not familiar with the data space. It’s worth seeking help from data folks in your network, or starting with a contractor to bootstrap the process.
Summary
Putting it all together, once you’ve read the above articles, my opinion is that you should:
- Get clear on the business needs that’d be addressed by a data person.
- Err on the side of not hiring prematurely – consider a contractor or rely on your current employees.
- When you’re ready to hire, sketch out a high-level plan for the person’s first 90-180-360 days.
- Run the plan and job description by some data people you trust.
- Possible title for the job ad: Data & Analytics Lead or Head of Data & Analytics (but you want a hands-on person, so make it clear that this is an individual contributor role initially).
- Make the plan a part of the job ad – it helps with aligning expectations.
- Ideally, get data people you trust to help you with the hiring process.
- Screen out specialists early, regardless of past titles and pedigree.
- Make expectations as clear as possible during the hiring process – especially if the person hasn’t worked with a startup before.
- Hire someone who’s a great fit who would help take your business to the next level.
Any thoughts or suggestions? Please contact me – I will make edits to this post based on feedback.
Public comments are closed, but I love hearing from readers. Feel free to contact me with your thoughts.