Narrow Wedge into a Legacy Market

Sources#

Summary#

John Glasgow's answer to the standard objection that you must be feature-complete to disrupt entrenched enterprise categories (ERP, CRM, HRIS). Campfire disproved it by being the best for a very narrow profile rather than broad: target tech companies that had outgrown QuickBooks but use only a fraction of NetSuite, solve a handful of features they actually need extremely well, and ignore the incumbent's vast unused surface. Within nine months this narrow focus was pulling customers off NetSuite — the supposedly mature, complex incumbent.

The "you don't need to be feature-complete" insight#

"Yes, the incumbents are incredibly deep and broad — but within our initial market of tech companies, they're actually not using a lot of NetSuite."

The leverage: incumbents are feature-complete for the whole market, but any specific segment uses a small subset. So the disruptor's job is to identify the high-value subset for a sharply-defined customer and beat the incumbent there. Campfire's chosen subset:

Approval workflows — "doesn't sound sexy, but a key reason folks need it for audit; QuickBooks doesn't have it."
Multi-entity accounting — subsidiaries; "in the AI-native world folks go multi-entity very quickly."

Define the customer narrowly, too#

The wedge is a customer profile as much as a feature set: not "anyone on QuickBooks" (that's the PLG/SMB/bookkeeper market, "a very different product to build") but the in-house accounting team at a tech company bringing accounting in-house. Glasgow chose pure sales over PLG precisely to serve this profile. "Own that one moment in time for tech companies." Narrowness compounds: a narrow feature set for a narrow customer at a narrow moment is what let four employees displace a NetSuite install at a $300M-revenue company.

Mockup-first MVP#

The wedge was validated with an almost-embarrassingly-thin first product: the first paying version was a Google Sheet iframed into a login, with Brex and Mercury feeds in a tab and a bunch of spreadsheet formulas — shipped <1 month into YC. "Very rudimentary," but real customers paid for it. The YC mantra ("how fast can you get paying customers? we don't want free") forces you to find the wedge empirically rather than build broad on spec. (Cf. Disposable Micro-Apps and Zero-Friction Scope Creep: ship the thinnest real thing, resist breadth.)

The wedge-flip lesson#

Campfire's wedge wasn't right on the first try. They applied to YC as a full general-ledger ERP, then briefly led with a revenue-recognition/SaaS-reporting add-on as the entry product — but "that cohort didn't convert well to the core product." So ~month 3 they flipped it: core ERP became the primary purchase, revenue an add-on. Lesson: the wedge is the entry point that converts to the core, not just the easiest thing to sell — test conversion, not just adoption.

The other wedge shape: software on top of someone else's hardware (DroneDeploy)#

Campfire's wedge is a narrow feature set for a narrow customer inside a software category. DroneDeploy (Kevin Spain, Emergence Capital, 2026-07-29, practitioner-opinion, investor-COI) is the same instinct pointed at a market where the incumbent surface is hardware: while most commercial drone companies built full-stack software-plus-hardware, DroneDeploy shipped software any drone maker could plug into — "it rode the industry's falling hardware costs instead of competing against large hardware makers." The wedge is a layer, not a segment: pick the part of the stack whose cost curve is falling under you and refuse to own the rest.

Two things this adds to the page, and one it doesn't.

The wedge's real cost was market education, not features. Spain's first "learning" is be an assembler: years spent with construction foremen and energy engineers who had never flown a drone, "showing them what a single flight could tell them about a job they already knew how to run." Campfire's wedge landed against a customer who already knew they needed accounting software; DroneDeploy's landed against customers with no category. Where the incumbent is a manual process rather than a product, the wedge is priced in years of teaching, and the first six-figure enterprise deals came ~2 years after a pre-revenue Series A.
A third exit for a wedge, alongside "become the incumbent" and "stay narrow". The Open Question below asks when narrow-but-best has to become the broad incumbent it displaced. DroneDeploy took neither branch: it stayed a layer, integrated deeply with the ecosystem — and was acquired for $845M by Procore, a partner long before it was an acquirer. Being the best-integrated layer in someone else's platform is a terminal state the wedge framing above doesn't enumerate.
What it is not evidence for. One outcome, narrated by the investor who funded it, with the losing full-stack competitors named and never examined. The software-first choice is asserted as the cause; nothing here rules out that it was a correlate of the teams that survived a decade.

Connections#

Compounding Data Moat — what the DroneDeploy wedge converted into: a decade of job-site imagery that was worthless until vision models could read it, then priced at exit
John Glasgow / Campfire — the source case
The AI-Native Safe-Choice Inversion — the macro tailwind; the wedge is how you land customers the inversion makes reachable
Problem-Solution Fit Discipline — narrow focus is problem-solution fit applied to GTM; founder-market-fit picks the right subset
Zero-Friction Scope Creep — the discipline that keeps the wedge narrow instead of sprawling into feature-completeness
Disposable Micro-Apps — the Google-Sheets MVP is the "thinnest real artifact" instinct
Product Velocity as Moat — once the wedge lands, velocity is what expands coverage without losing the customer
AI-Native Startup Lifecycle — the Idea/MVP-stage discipline of the Founder's Playbook, executed
Vertical Slice Tracer Bullets — a narrow feature set that fully solves one profile is a vertical slice, not a horizontal layer
The 1% Rule for Wedge Selection — the other axis a wedge has to clear: this page picks against an incumbent product's unused surface, Dean's rule picks against the frontier model's capability curve (build where it succeeds 0–1% of the time, not 20%)

Open Questions#

A wedge works going in; does it constrain going out? Campfire now serves public companies — at what point does "narrow-but-best" require becoming the broad incumbent it displaced, re-incurring NetSuite's complexity?
The wedge-flip shows the first wedge can be wrong. What's the fastest signal that a wedge converts to the core vs. merely sells — Campfire took ~3 months; can it be read sooner?

Sources#

The ERP for the AI Revolution is here
DroneDeploy Didn't Win on Hardware. That's Why Procore Just Paid $845 Million. — Kevin Spain, Emergence Capital, 2026-07-29 (practitioner-opinion, COI: led the Series A, board seat): "Software First" and the "be an assembler" learning — the layer-not-segment wedge, market education as the wedge's cost, and Procore as partner-then-acquirer