Building a Proprietary Pipeline When You Can't Afford PitchBook

PitchBook, Grata, and Sourcescrub are built for tech and healthcare M&A. They won't help you find a $2M plumbing company in Charlotte. The lower middle market of boring businesses runs on public data that most buyers never systematically collect.

Free and low-cost data sources that actually work

• Google Maps / Business Profiles, the single best source for local service operators.
• State licensing boards, HVAC, plumbing, electrical often have public licensee databases.
• Secretary of state filings, LLC/Corp registrations with registered agent info.
• YellowPages and industry directories, still relevant for owner-operated trades.
• WHOIS and domain records, age, privacy protection, registrant org (when not redacted).

The hard part isn't finding the data. It's deduplicating, enriching, and keeping it fresh. A plumbing company listed on Maps, YP, and a state license database is three records that should be one.

The proprietary moat is freshness + signals

A static list decays at 15–20% per year (businesses close, rebrand, sell). A living database that re-crawls weekly and scores owner neglect gives you leads your competitors don't have, because they're still working from a CSV they bought in 2023.

Building the universe without a data budget

Start with licensing boards for your trade, export or scrape where legal, normalize company names, match to Maps listings. Add secretary of state entities for registered agents and formation dates. Layer review counts and website status as size and neglect proxies. Document every merge rule so when an intern leaves, the pipeline does not reset to zero.

• Define 'in market' precisely: MSA, county list, or 45-minute drive radius from platform HQ.
• Tag records with source and last-verified date; stale phones waste searcher credibility.
• Segment by estimated headcount proxy: review velocity, truck photos on GMB, job postings.

When to upgrade to paid tools

After your first close, invest in: a dedicated CRM (Affinity or HubSpot), a QoE firm on retainer, and a part-time analyst to manage outreach. Before your first close, every dollar should go toward conversations with owners, not software subscriptions you won't use.

Outreach copy that respects owner pride

Owners built businesses when websites did not matter. Praise operational reputation ('four-star service for twenty years') before mentioning digital gaps. Searchers who lead with 'your marketing is broken' get hung up on; searchers who lead with succession and employee continuity get meetings.

PitchBook-shaped questions, local answers

LPs trained on tech M&A will ask about TAM and fragmentation. Answer with trade-specific math: licensed HVAC contractors in your state, average trucks per $400K SDE shop, owner age cohort retiring over the next decade. Public data plus systematic collection beats a $40K data subscription that does not list the plumber you are about to close.

Data hygiene as competitive moat

Your moat is not the first scrape, it is the fifth refresh. Assign dedupe keys (phone, address, license number), track merged records, and archive businesses that consolidated into PE platforms. Searchers who reuse the same CSV for 18 months wonder why response rates collapse; living data explains which owners flipped from Tier 3 to Tier 1 when their web domain expired.

Budget 4–6 hours per week per metro in thesis for maintenance: new listings from Maps, license board deltas, and neglect re-score. That is cheaper than one dead LOI month and more defensible in LP Q&A than claiming a 'network effect' you cannot document.

Analyst playbook for enrichment

Hire or intern analysts with clear SOPs: verify phone with two sources, mark website status with screenshot date, flag franchises automatically. Quality beats quantity, 2,000 verified records outperform 8,000 scraped rows with 30% wrong numbers. Wrong numbers burn searcher reputation with owners who tell their wholesaler friends.

• Weekly QA sample: 50 random records called or spot-checked.
• Escalate PE-owned brands to separate list, different outreach thesis.
• Export CRM snapshots for LP appendix, proves systematic work.

Self-funded vs. traditional sourcing load

Self-funded searchers often run leaner, one or two people, no analyst bench, which makes infrastructure choices more critical. Automate enrichment where possible, ruthlessly narrow geography, and say no to side projects until LOI. Traditional funds with salary runway should still cap metros at two until meeting velocity proves the model.

Both models win when the database is living and the CRM is honest. Both lose when the searcher confuses attending conferences with contacting owners.

Building vs. buying data

Scraping Maps yourself is cheap until you account for engineer or analyst time, dedupe bugs, and stale phones. Buying a living dataset purpose-built for boring-business acquisition trades CapEx for speed, often correct before your first close when searcher hours should be on owner calls. After close, some funds insource maintenance; others keep external refresh and focus corp dev on integration.

Bookmark state licensing portals and set calendar reminders for quarterly exports, license data catches new entrants and inactive status faster than Maps alone for plumbing and HVAC universes.

Pair license exports with your CRM dedupe rules the same week, stale universe data is the silent killer of search fund meeting rates in year two.

Review pass reasons monthly, if owners cite 'not selling' but accept coffee, your copy may be too transactional; soften succession framing before you blame list quality.