I was reading the latest episode of the AE roundup which had a reflection on Data Council pre and post covid from Drew Banin:
A lot has changed since the last Data Council. If I was in a room with folks like Claire, Emilie, Anna, Scott, Taylor, Max, and Julien in 2019, I’d be in a room full of practitioners. Today though, that same group is now composed of people building or investing in data tools. I don’t think I had internalized exactly how big or efficient the data-expert-to-data-startup pipeline had become, but it was hard to miss it at the event this week. I guess the thing that feels resonant to me upon reflection is: Data Council changed in part because we changed, and it’s helpful for me to keep that in mind.
What is behind the trend of practitioners becoming founders at Modern Data Stack companies?
Here is my take and a few ideas:
We’ve gone from organisation to organisation and solved the same problems. It’s fulfilling to a point, but it doesn’t scale. There will only be so many times that a practitioner will want take a new org from zero to one.
There is a huge shortage of experienced practitioners
There is inertia and confusion in companies around data strategy: many follow the “let’s hire a VP/Head of Data and it’ll all be ok” approach
Practitioners have a huge amount of distraction to deal with:
making business cases
hiring, leading and maintaining teams
dealing with organisational problems
upstream issues which aren’t a priority to be solved by PMs
The orgs that attract the talent may end up with data functioning well, but the others remain a mess which one day we’ll have to clean up should we want to move on. We ask ourselves whether this needs to be the case?
Some of us have founded consultancies as a response:
Allowing us to scale from a handful of orgs to tens in a career.
Divorcing us from organisational problems, confusion and inertia to a level… if they’ve come to the point of asking for help, then they realise and are invested in a solution they are willing to pay a lot for.
It allows for ownership of our endeavours in a better way than joining an org internally where we may not hold any options, and may not have any power to drive change. The “how” is the product for these consultancies, and having this the absolute best it can be, regardless of what an internal enterprise architect might like.
This is occurring inside VCs in a slightly different form, where they are building data orgs to serve their early portfolio companies who can’t afford to have their own data people, thus increasing the likelihood of success.
There is a huge opportunity in front of us as practitioners, as there are gaps in the modern data stack despite the unbundling debate; there is no-one better placed to identify and exploit these gaps than a practitioner.
The modern data stack is reaching a critical state of maturity, with many organisations having good foundations in place. However, we have yet to fully exploit the potential of this higher data maturity. Practitioners, when not fire-fighting and doing engineering, are some of the only people who know what good exploitation of a mature data stack looks like.
Disciplines like Analytics Engineering are leading us towards purpleshift; we’re now more technically able than ever before, with many of us delving into software engineering, platform and architecture. We have also become more focused on solving problems for our stakeholders, with a product mindset, so why not for all stakeholders in the same manner?
We are better connected than ever before - we have support from large, growing communities of like-minded people:
We’re giving each other the confidence to try something new
We can find problems to solve and products to build from our own experiences and those of our communities
We can adopt product-led growth because of this
Data people have a good view of how a whole business works
I’ve said this before: I do believe data people are usually more commercially-minded than engineers, and are more technically-minded than wholly commercially-focused teams like marketing, sales and finance
We sit in the middle of everything
Therefore, we know what a good working model of a business looks like, whether in microcosms or at a macro-level
We really believe we can solve a new problem or an existing one in a better way than before, and we’re making whole categories of products not thought of before. Monte Carlo and Avo are great examples of where founders made the category to tackle difficult unsolved problems.
We have a purity of purpose in fulfilling known use cases and stakeholder need, as data practitioners, that software engineers and product managers at FAANG companies may not have:
This is why you see some projects come from FAANG as OSS and aren’t really ready to be deployed in most businesses to solve stakeholder need
It’s also why you see some products fall short of the mark when offered by the cloud vendors - they’re divorced from their target market in a way a good data practitioner will never be
We aren’t hindered by serving the overall goal of the cloud vendor, which will distract time, resource and focus away from a product in making it the best in its class. We can be single-minded in solving the problem we’re focused on, or delivering the experience we want in the very best way we can.
The shortage of data practitioners means that there is little to no risk from trying to found companies, save potentially temporarily lower pay. Employers can’t be fussy about practitioners having attempted to found, leading to a less structured CV. I’d even go the other way, in seeing candidates who have tried to found as more interesting if you need highly proactive and autonomous hires.
As more practitioners leave the market as employees to found, there are fewer to hire to build tools and stacks internally, leading to more need for the products these new founders are building. It's almost a self-fulfilling prophecy that many of these data companies will be successful (especially relative to startups not building data SaaS tooling). Internal data teams are accustomed to buying products to compose their stack from - they don’t have the hangups of some software engineering teams about having a “buy nothing build everything” culture (this is not always a bad thing), and buying from another practitioner who understands what you’ve been through is a compelling experience.
This shift towards using the companies the practitioners have founded vs having them internally may not be a bad thing; building some of these systems internally carries a very high total cost of ownership when you take into account time to build, maintenance and opportunity cost of both. Founders want to operate at scale and this comes with efficiency not possible with a one-time solution.
VCs are keen to fund:
They believe the market for data products is orders of magnitude bigger than it is now.
They have funds to deploy, that they need to deploy to get return on.
The market may seem choppy at the moment but that makes it a good time to invest in pre-seed, seed and series A orgs. And even one or two later stage ones.