Home Industries Vertical SaaS
Industry

Vertical SaaS — niche software the horizontal databases file wrong

The software industry's data infrastructure was built to track funding events.

Bootstrapped vertical SaaS — shop-management systems, lab compliance tools, niche field-service platforms — generates none, so the databases organized around rounds and investors never see it.

For buyers consolidating vertical software, that unfunded long tail is not a rounding error. It is the thesis.

100M+
classified domains screened
700+
industry categories resolved
300+
organisations served
0
funding events required to be found

The funding-radar blind spot

Bootstrapped vertical SaaS lives entirely off the funding-indexed map — and it is precisely the population most worth buying. A full-web census finds it the way its customers do: by reading what the product says it does, for whom, at what price.

Invisible by design

No round raised, no press release, no database row. Bootstrapped products are fully documented on their websites and completely unindexed in funding-centric tools.

ICP in plain language

Each product publishes its niche, pricing tiers, integration partners, and customer logos — because narrow targeting is the marketing strategy. The evidence is right on the site.

Artisanal methods miss

Conference lists, marketplace crawls, and referral chains each start from a partial index of the niche. They find real companies and miss unknowably many more.

Census inverts the logic

100M+ classified domains across 700+ categories. The software population falls out of the classification regardless of funding history, press, or directory presence.

Examples hiding in plain sight: a scheduling product for equipment rental fleets, a compliance platform for calibration labs, a quoting engine for custom fabricators — each fully documented on the web, none in any funding database.

Reading a product the way its buyer reads it

Two passes cut through software's noisy vocabulary. Pass one triages cheaply; pass two reads each site the way a customer and then an acquirer would.

Pass 1 — triage

Live product or dead domain? Actual SaaS or agency with a product page? In-scope vertical or generic tool? Disposed cheaply at scale.

Pass 2 — deep extraction

Full LLM read against your written thesis. Each extraction lands as a quoted snippet with source URL. Fifteen signals, three scores.

Five layers the deep pass reads

ICP evidence

Who the product serves, in its own words — corroborated by customer logos, testimonials, and case studies from inside the niche.

Commercial motion

Published pricing structure (per-seat, per-location, per-transaction), trial mechanics, self-serve vs. demo-gated enterprise flows.

Moat furniture

Integration pages, API docs, marketplace presences — the partner ecosystem that makes a vertical product sticky.

Delivery model

True multi-tenant SaaS vs. installed software vs. services wearing a product costume — a distinction acquirers care about intensely.

People & pulse

Founder visibility, team pages, hiring posture, changelog cadence — the liveness check that separates maintained products from frozen sites.

Mandate Fit
70%
Outreach Suitability
20%
Transition Context
10%
Group ownership zeroes the total. ICP re-runs against revised theses included throughout.

Six signals that sort vertical software

From the standing 15-signal framework; these six carry vertical-SaaS theses. Every extraction is quoted and sourced.

Vertical specialization & documented end-market exposure

The load-bearing signal: which industry the product serves, evidenced from ICP language, customer walls, and case studies — not a category tag. We extract at the granularity the product itself claims, then corroborate; “software for independent pharmacies” is a thesis-relevant fact, “healthcare IT” is filing noise.

Digital-commercial maturity

Published pricing, self-serve signup, trial mechanics, billing language — the commercial architecture is on the website and discriminates sharply. Per-seat pricing with a self-serve motion implies one kind of business; “contact sales” implies another; an invoice-shaped services page implies a third. We capture the motion as evidence because it predicts revenue quality and post-close build requirements.

Partner & channel ecosystem position

Integration ecosystems are vertical SaaS's moat made visible: the accounting packages, industry hardware, and platform APIs a product connects to define its workflow position. Partner pages and API docs are read and quoted — a product woven into its vertical's stack churns and prices differently than a standalone tool.

Founder-led / family-led association

Bootstrapped SaaS is founder-context-rich by nature — bylines on changelogs, origin stories rooted in the industry, “built by a contractor for contractors.” Captured strictly as published; combined with operating history and independence evidence, it feeds Transition Context without inferred personal circumstances.

Website/news activity trajectory

A live product leaves footprints — dated changelog entries, release notes, blog cadence — while a zombie leaves a beautiful site frozen in 2021. For software acquirers, the distinction between maintained and abandoned is worth more than most firmographic fields combined. Captured as dated evidence, not impressions.

Management professionalization & ownership

Consolidator branding in the footer, “family of products” pages, acquisition announcements — each zeroes the score with language quoted. Roughly one in ten keyword-perfect candidates fails on ownership; software's version is the quietly-acquired product still wearing its founder-era site years later.

A worked example: software for a trade, screened

Composite reflecting real run behavior. Thesis: bootstrapped SaaS serving industrial and trade businesses — shop management, field service, quoting, compliance — North American, product-led, independent.

Census discoveryconfirmed fit

Job-management product built by a former shop owner, self-funded eleven years, per-seat pricing published, 400 logos from one trade, founder still writing release notes. No round, no press, no database row anywhere — surfaces only from full-web coverage.

Keyword-missed fittrade language

Product describes itself as “dispatch board for service contractors” — trade language, not software-category language. Keyword tools miss it entirely; the census classified it from ICP evidence on the product's own pages.

Insufficient evidenceflagged honestly

Products behind login walls with one-page marketing sites, tools whose sites cannot support classification either way — labeled honestly, never force-classified.

Three exclusion patterns the screen catches

Services in disguise

Agency whose “platform” page fronts a services business — classified out on delivery-model evidence, quoted.

Zombie product

Polished site, changelog stopped two years ago, careers page empty — activity-trajectory evidence, documented.

Already acquired

Footer announces membership in a software group's “family of brands” — ownership zeroed, sentence quoted.

Eleven published specimens from industrial verticals demonstrate this discipline: view the sample page.

What this screen cannot tell you about software

Every sourcing method has limits; ours are stated per record so the rest of the deliverable earns trust.

No financial metrics

ARR, churn, NRR, seat counts, margins — none are published and we do not manufacture estimates from page text.

No technical diligence

Code quality, architecture, and the single-tenant skeletons in a “cloud” product's closet belong to technical review.

No intent claims

No website signal supports willingness-to-transact — we categorically refuse to sell that inference.

Thin-site honest flags

Enterprise motions with sparse public sites get flagged as insufficient evidence rather than guessed.

What it does instead: complete coverage, delivery-model and ICP classification with quotes, commercial-motion and ecosystem evidence, liveness signals, ownership screening — each with a source URL. The screen compresses discovery months into weeks; the data room is still the data room.

From census file to term sheet conversation

1

Universe in

Scored CSV plus evidence appendix — served vertical, delivery model, pricing motion, ecosystem, liveness, ownership — quotes and URLs per row.

2

Thesis out loud

You disagree with rankings; we translate the disagreement into ICP terms and re-run the scored universe. Included, and usually the moment the thesis gets sharp.

3

Founder-grade outreach

Letters that cite the product's actual integrations and niche read as customer-level attention. Founders who ignore brokers answer those.

4

Watch the niche

Annual monitoring re-reads the universe: new pricing pages, stalled changelogs, new footers announcing new owners. Vertical software moves; the map moves with it.

€4,900
proof project
€9,900
full universe + shortlist
€18,000
annual monitoring / thesis

Custom ICP re-runs included throughout. 300+ organisations served, including a leading European telecom operator and a major airline-metasearch platform.

Funding-indexed databases and the census, compared

The claim in one sentence, varied everywhere on this site: leading company databases index the companies they found; we start from the entire active web and read it against your exact thesis, with evidence both ways.

QuestionFunding-indexed / horizontal databasesFull-web census screening
When does a company become visible?When capital or press touches itWhen its website exists
Bootstrapped niche leader, 12 years oldUsually absentRead, classified, scored — with its pricing and ecosystem quoted
Product described in trade language, not software languageMiscategorized or missingClassified from ICP evidence; the keyword-missed cohort is a standing specimen section
Quietly acquired product with founder-era siteListed as independentOwnership zero, footer language quoted
Services firm with a “platform” pageListed as softwareClassified out on delivery-model evidence, documented
Category granularity“Software — vertical: healthcare”The product's own ICP claim, quoted, corroborated
Thesis iterationNew filters on the same indexFull re-run against your revised ICP, included

Where vertical software hides, surface by surface

Each artisanal discovery surface is a legitimate lens — the failure is treating any of them, or all stitched together, as a census.

App marketplaces

Surface only products that chose a platform's ecosystem. Excellent coverage of one stack's satellites; silence about everything built outside it.

Conferences & vendor lists

Capture founders who buy booths — a marketing-budget filter, not a quality filter. The most efficient bootstrappers often skip them deliberately.

Review platforms

Index what users bothered to review. Over-weights horizontal tools with large seat counts; under-weights the twelve-person product running 400 machine shops with zero public reviews.

Diminishing returns

Artisanal surfaces overlap heavily — working more yields sharply diminishing discoveries as the same funded, marketed, reviewed products keep reappearing.

Census advantage

Census coverage does not diminish — its marginal discoveries are precisely the products no surface carried. For consolidators whose edge depends on meeting founders before the auction, that unshared cohort is the entire strategic value.

Frequently asked — vertical SaaS screens

By not starting from databases. The classified universe covers 100M+ active domains sorted into 700+ categories; a software product with a live website is in it regardless of funding history, press coverage, or directory presence. The screen then reads each candidate site's ICP language, pricing, and integrations directly. Visibility to us requires exactly one thing — a website — which is the one thing every operating SaaS business has.

That distinction is a core classification, made on evidence: multi-tenant product language, self-serve or trial motions, release notes, documentation depth, versus proposal-shaped engagement language and portfolio pages. Hybrids — productized services, single-tenant deployments — are labeled as what the evidence shows, with quotes, and your ICP decides their fate. The point is that the label is auditable, not asserted.

One engagement, several ICP definitions. The universe build is the expensive step; scoring against multiple written theses — or re-scoring as they evolve — runs on the same scored population and is included. Buyers consolidating adjacent verticals typically run one census and iterate ICPs per niche, which is dramatically cheaper than serial artisanal searches.

No. Neither is website-visible, and vendors who estimate them from traffic or page text are guessing with confidence intervals they do not disclose. We report what is evidenced: pricing structure, plan tiers, customer-logo density from inside the niche, hiring posture, activity cadence — quoted and sourced. Those bound the commercial reality honestly; the data room provides the rest.

The discipline, which transfers exactly: 8 top fits, 5 keyword-missed fits, 5 documented exclusions, 2 insufficient-evidence flags, every classification justified with quoted site text and every exclusion explained. Software screens produce the same artifact against a software thesis. Request a specimen via the sample page and read the exclusion log first — it is the fastest way to judge whether the evidence standard is real.

Adjacent reading

Managed IT servicesERP & IT consultanciesSpecialty wholesale distributionMethod: two passes, 15 signalsPricing, publicAll industries
What we refuse to sell: no “ready to sell” flags, no revenue or EBITDA guesses, no owner-age profiling, no distress detection — and no engagements in consumer-captive verticals. Read our standards; serious buyers tell us this page is why they trusted the rest.

Map a software niche the funding radar never saw

Name the vertical and the thesis. A specimen report arrives the same day; a scoped proposal for the niche follows.

Request the specimen report