Fig. 001 — Breadcrumb, As Built

Find every buyer in your market. Then sell to them from the same system.

Breadcrumb maps a client's entire industry: 17,100+ companies, their plants and decision-makers, researched with cited sources and scored A–D against their ideal customer. It's now the CRM their sales team works deals in. I build the same system for your market, around your sources, your buyer and your sales process.

Breadcrumb company list showing 2,675 matching companies ranked by score with tier badges, each best contact's job title and verified-email count, and hiring, news, and client signals, beside filters for AI tier, rules tier, signals, and contactability. Company names, websites, locations, plant and location counts, and contact names blurred.
// The companies list: 2,675 qualified leads out of 17,100+ companies, each with its tier, best contact and live signals. (Company names, locations and contact names blurred.)
ProjectBreadcrumb
ClientWithheld
EngineerD. Holdiman
StatusIn production
Scale1:1
Sheet01 of 05
17,100+ Companies Tracked
2,675 Qualified Leads
37,000+ Contacts Found 5,600+ with verified email
23,500+ Building Footprints Measured

Every company in your market, including the ones no list sells.

Bought lists are sorted by legal entity and out of date the day they arrive. Breadcrumb builds the market from 15+ sources and keeps the evidence behind every record.

Interactive North America pin map in the Breadcrumb platform plotting over 3,000 company headquarters and plants, clustered by region and filterable by lead tier.
// 3,000+ headquarters and plants plotted and clustered: the whole addressable market on one map.
Sources

Directories, publications, maps and registries

Directories, OCR'd trade publications, map data and enrichment APIs swept state by state. For your market, also the places your buyers show up: permits, procurement, customs and property records.

Dedup

One record per company

97,000+ sightings became 17,100+ companies. Merging keeps the evidence, so every field traces back to where it was seen.

Structure

Who owns whom

Parents, sister companies, brands and historic names mapped, so a subsidiary stops hiding behind its holding group.

Facilities

Every plant on a map

23,500+ building footprints measured for square footage, a proxy for capacity and budget, and plotted for territory planning.

A researched dossier on every company, with the source for each claim.

An AI research pass reads each company's site and public footprint. A scoring model then ranks it against your ideal customer.

Company dossier in the Breadcrumb platform showing an existing-client badge, revenue estimate, segment, industry codes, parent company, a panel for the company's deals, a facility chat with suggested questions, and a score-history timeline. Company name, website, address, revenue, plant count, and parent company blurred.
// A company dossier: revenue estimate, industry codes and parent company, with deals, a facility chat and score history on one page. (Company name, address, revenue, plant count and parent company blurred.)
AI Research

Cited dossiers

What they make, where, who owns them, a revenue estimate, buying signals, and which of your products fit which of their operations. Each claim is stored with the URL it was read from.

Scoring

Tiers A–D on your criteria

Product fit, size, facility scale and contactability, and every score explains itself.

Contacts

Decision-makers, credit-safe

Plant managers, operations VPs and owners, pulled only for companies in qualified tiers, then validated.

Who to call today, and why.

Research jobs re-run on a schedule or on demand, and the call queue re-ranks as the market moves.

Breadcrumb prospecting queue on the Today page ranking 196 companies worth a call, each card showing its tier, hiring and call chips, and the best contact's job title, with company names, contact names, email addresses, and headlines blurred.
// The Hot List on a rep's Today page: ranked, with the best contact and the latest headline for each company. (Company names, people, contact details and headlines blurred.)
News

News sweeps

Expansions, acquisitions, launches and leadership changes, classified as growth or risk and attached to the company.

Hiring

Hiring signals

A company hiring plant managers or line operators is investing in capacity. Hiring flags move it up the queue.

Prioritization

The Hot List

News, hiring and tier blended into one ranked queue on each rep's Today page, with the reason, the best contact and the latest headline.

Data Hygiene

Data that stays clean

Dead domains flagged and repaired. Duplicates merged with a preview first, then re-scored.

Then it became the CRM the sales team works in.

The same records carry each deal from first call to active customer, with mail, meetings and quotes in one place.

Pipeline board in the Breadcrumb platform showing the Leads lane with Identified, Working, and Engaged columns, each headed by its move-on rule, and cards showing rank, days in stage, off-track and open-commitment flags, and the next task with its due date. Company names, people, task text, and dollar values blurred.
// The pipeline board: every card has a next task, and every column states its move-on rule. (Company names, people, task text and dollar values blurred.)
Pipeline

A pipeline with written rules

Each deal has an owner, a stage and one open next task. Each stage has a written exit rule.

Mail & Meetings

Mail and meetings on the deal

Microsoft 365 mail and Teams transcripts are filed on the right deal. Reps send from their own mailbox, and a person approves every send.

AI Brief

A sourced deal brief

Built from facts read from the deal's mail, documents and transcripts. Reps can correct or pin a fact.

Quotes & Reports

Quotes and reports

Quote documents generated from templates, with revisions kept, and a weekly pipeline overview for leadership.

Claude Connector

Work it from Claude

A connector with 35+ tools for the morning brief, call prep and wrap-up. Each person signs in as themselves, and every write is logged.

Rollout

AI proposes, a person decides

New AI behaviours run in shadow mode first.

Your market, your buyer, your sales process.

Nothing in Breadcrumb's machinery knows about food production. A few layers change for each client, and the rest carries over.

Changes for you

  • Sources
  • Ideal customer profile
  • Fit logic
  • CRM stages and rules
  • Your current CRM's history, imported

Carries over

  • Discovery
  • Dedup with provenance
  • Cited research
  • Structure mapping
  • Facility research
  • Contact discovery
  • Freshness jobs
  • The CRM

The same system could map logistics operators for a warehouse-automation vendor, medical practices for a device company, or contractors for a building-materials brand.

From survey to as-built.

01

Survey

I learn who you sell to, where those buyers show up, and how your team works a deal today.

02

Draft

You get a drawing of your system: the sources, the scoring rules, the CRM stages, and what AI does versus what a person approves.

03

Build

Discovery, research, scoring and the CRM, built for production and running on your data.

04

Prove

Your reps work real leads in it, alongside the old way, until the numbers say it's better.

05

Operate

It keeps running, and it keeps getting extended as your sales process changes.

Why one good engineer beats the usual options.

You've got three ways to get a system that finds and works your buyers. Two of them are how it usually goes wrong.

Option A — The agency

You brief a salesperson. A junior builds it.

  • Account managers between you and the code
  • Scope written before anyone understands the workflow
  • Handoffs, hourly burn, and a maintenance retainer
Option B — Another SaaS tool

You buy a list and rent a CRM.

  • The list is sorted by legal entity and goes stale
  • The CRM doesn't know your market
  • Per-seat pricing, forever, for a partial fit

Engineering deep-dives on what I actually build.

Long-form write-ups on production scraping, browser automation, AI extraction pipelines, and the back-office systems that hold them together.

Python Web Scraping

Mining Buying Signals from Public Exhaust — Jobs, News, and Import Records

How to detect purchase intent from what companies do in public — job postings across 12 ATS platforms, LLM-classified news monitoring, customs import records, and domain-health checks — and why signals should rank your call queue but never touch your lead score.

Read the article →
Python Web Scraping

Entity Resolution for Scraped Company Data — One Company, Every Source, Zero Lost Evidence

How to deduplicate company records scraped from dozens of sources into one canonical row per real-world company — the match ladder, fill-blanks-only merging, provenance rows that never get destroyed, and the counting bugs that bite everyone who builds one of these.

Read the article →
AI Data Extraction

The Contact Enrichment Waterfall — Verified Decision-Makers Without Burning Credits

A production architecture for B2B contact enrichment — free website extraction first, credit-gated API reveals second, premium sources last, with score gating, no-double-spend guards, and upward-only verification status so every credit lands on a company worth calling.

Read the article →

Tell me who you sell to. I'll scope the system that finds them.

Send the market you sell into and how your team works deals today. I'll reply with what your Breadcrumb would include and what it would take.

No pitch decks, no sales calls — just a straight answer on what should be built and what it would take.