Breadth-Then-Depth Platform Strategy
Go wide to discover your niche, then put all your wood behind the few arrows that lift the rest
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 93%
The standard advice is to do one thing brilliantly. GitLab did the opposite: it went deliberately broad across the entire software development lifecycle to earn platform status, and only then — in 2023 — consciously pivoted to depth in a handful of key areas. The insight is that breadth is a discovery mechanism (you're finding your niche and your differentiation), and depth is the exploitation phase you enter once you've found it. Areas you leave shallow ride a rising tide from the deep ones.
Origin
GitLab's product strategy as described by CPO David DeSanto, with the pivot logic explicitly grounded in Geoffrey Moore's Crossing the Chasm — DeSanto's test for founders asking when to pivot is 'have you found your fit in the market?', after which you put all your wood behind that arrow. He notes HubSpot (Dharmesh Shah) independently ran a breadth-over-depth strategy too.
Core principles
- 01Breadth is exploration: when you're going really wide, you're trying to find your niche and what you can differentiate on.
- 02Depth is exploitation: once you've found product-market fit, put all your wood behind that arrow.
- 03Choose deep areas by leverage — the ones that most accelerate the customer's core job.
- 04Deep areas bridge to each other; adjacent shallow areas get a rising-tide lift and only need to be good enough.
- 05'Good enough' can be delivered by a lightweight connector to someone else's deep product rather than by building it.
- 06The dial is never permanently set — you swing back to breadth whenever you expand into a new market or SAM.
How to run it
- 1
Go wide across the customer's full lifecycle
Build presence across every stage of the customer's workflow to earn platform status. GitLab went from source code management into CI/CD, security, compliance, enterprise agile planning, business continuity, OKRs, monitoring, service desk — touching the whole SDLC.
Watch out Breadth is expensive and only defensible if you are genuinely aiming at a platform position, not collecting features.
- 2
Use the breadth phase as market research
Watch which of the wide bets produce real differentiation and real customer acceleration. The wide phase's job is to reveal the niche, not to win every category.
Pro tip Track which surfaces customers pull you deeper into unprompted — those are your depth candidates.
- 3
Apply the Crossing the Chasm test before pivoting
DeSanto's question to founders asking when to make the pivot: have you found your fit in the market? If yes, stop spreading and concentrate.
Pro tip Pair with Essentialism — DeSanto pairs Moore's book with Greg McKeown's Essentialism plus Moore's critical/core/context framework to decide what is critical, what is core expectation, and what you can say no to at low risk.
Watch out Pivoting to depth before you have fit just makes you deeply committed to the wrong thing.
- 4
Name the small set of areas that go deep
GitLab named five: source code management (and code review, IDE experience, remote development), CI/CD, security and governance, planning, and AI. Everything else is explicitly not going deep.
Pro tip The list must be short enough to be a real constraint. Five out of a whole-SDLC platform is a genuine choice.
- 5
Let the shallow areas be lifted, or connected
Areas adjacent to the deep ones get the rising-tide effect and only need to be good enough. Where a whole adjacent category is needed, ship a lightweight connector instead of a product — GitLab went deep on MLOps and covers DataOps with an integration to external platforms.
Pro tip A good-enough connector to a category leader is often a better customer outcome than a mediocre in-house build.
Watch out Good enough still has to be good enough. A shallow area that actively breaks the workflow undoes the platform argument.
- 6
Swing back to breadth for new markets
Breadth-over-depth remains the mode for new investment areas — GitLab currently runs it inside its AI strategy — and for any push to expand the serviceable addressable market.
In the wild
From 2019 GitLab pursued breadth over depth to build out a true DevOps platform across the whole SDLC. Having become the recognized DevOps platform leader (per user reviews and analysts like Gartner and Forrester), they made a conscious decision last year to pivot to depth over breadth in five areas that most accelerate software delivery.
→ The five deep areas bridge to each other and lift the surrounding shallow areas via a rising-tide effect, letting GitLab hold platform breadth without needing to be best-in-class in every stage.
GitLab started adding model-ops functionality in 2020 and invested heavily in MLOps — one of three pillars of that space. They judged DataOps would eventually be needed, but chose a lightweight connector to external DataOps platforms instead of building it.
→ Customers can do their full MLOps work inside GitLab successfully, with the DataOps gap covered by integration rather than by a shallow in-house build.
Common mistakes
Going deep before you know where your niche is
Depth is the exploitation phase. Committing all your resources to one area before the market has told you where you differentiate is a bet with no information behind it — the breadth phase exists precisely to generate that information.
Trying to be deep everywhere on a broad platform
It's unaffordable and unnecessary. The rising-tide logic says that if the leverage areas are genuinely deep and bridge to each other, the surrounding areas only need to clear a good-enough bar.
Treating breadth vs depth as a permanent company identity
DeSanto is explicit that breadth-over-depth hasn't left GitLab — it's the mode they use for new investments and new markets. The dial is per-bet, not per-company.
Is it for you?
Best for
Founders and product leaders building a platform play in a market with many adjacent workflow stages, deciding whether to widen or concentrate.
Not ideal for
Capital-constrained startups pre-product-market-fit, where breadth is unaffordable and the correct move is the standard one-thing-brilliantly advice.
From the transcript
“we were very much focused on a breath over depth”
“we now know there's key areas where we need to be really deep to help companies accelerate delivering software”
“we know if those are really deep the things that aren't as deep around them we'll get that like a rising tide LIF all bods…”
“I think when you're going really wide you're trying to find your Niche and like what you're going to be really good at and what…”
From the episode
The GitLab way: Kindness, transparency, and short toes
David DeSanto (CPO)