Articles & Case Studies
PipeRich Gets 2x More Qualified Accounts
PipeRich, an AI outbound agency in Waterloo, Ontario, uses DiscoLike as the primary account discovery source for 56 clients. Over two months it supplied 2.8 million net-new company domains across roughly 720 campaigns, and it yields roughly 2x more qualified accounts per exported account than the next-best source PipeRich runs alongside it.
Client Overview
- Industry: AI-Driven B2B Outbound Sales
- Company: PipeRich
- Use Case: Primary account discovery for multi-client outbound campaigns
- DiscoLike Features Used: Lookalike Search, Natural Language ICP Search, Discover MCP, API, Free Count
Executive Summary
PipeRich runs outbound on behalf of B2B clients. Its AI writes a personalized message for each target company, submits it through the company’s website contact form, and relays interested replies back to the client. Every campaign starts with the same question: which companies should we contact?
PipeRich has answered that question with DiscoLike since mid-2025. In the two months from late July to late September 2026, DiscoLike supplied roughly 2.8 million net-new company domains, feeding about 720 campaigns across 56 clients where it is the primary source. Measured on qualified accounts per account exported, DiscoLike leads the several other data sources PipeRich runs alongside it by roughly 2x, with the widest gap in the hard-to-find SMB verticals many of PipeRich’s clients sell into.
“Overall, my favorite data source, and we love leveraging it on behalf of our clients.”
Jeff Rolf von den Baumen, Co-Founder & CEO, PipeRich
The Challenge: Every Client Is a New Market
An outbound agency does not get to pick one ICP and optimize it for years. PipeRich’s clients sell into MedTech, HVACR, health plans, construction, commercial real estate, insurance agencies, mortgage brokers, and dozens of other verticals. Most of those markets are made of small and mid-sized businesses that rarely maintain a LinkedIn company page, which means the usual databases either miss them or list them under a generic industry tag that hides what they actually do.
PipeRich’s model makes list quality unusually visible. Each account gets a personalized message and a form submission, so a bad account is not a bounced email. It is wasted work and a wasted first impression. PipeRich validates every account again after export, so the number that matters is not how many domains a source returns. It is how many survive validation.
Two constraints follow from that:
- Coverage in obscure verticals. If the source cannot see a market, no amount of filtering fixes it.
- Yield per exported account. Every export costs validation time, so a source that returns more junk costs more than it looks.
The Solution: Website-Sourced Discovery for Every Segment
DiscoLike indexes over 80 million business websites and matches companies by what their sites say, not by self-reported taxonomy. For PipeRich, that turned the same workflow into a repeatable process across very different clients.
1. Segment the market by business model
PipeRich has Claude work through the Discover MCP and API to find parts of a client’s market that have not been targeted yet. Claude splits a market into single business-model segments, one type of business per campaign, so the messaging and the list stay aligned.
2. Find lookalikes from the client’s best-fit examples
Each segment starts with the client’s best existing customers as seed domains. DiscoLike’s lookalike search finds companies whose websites read like those seeds, which surfaces the niche and small-business accounts that industry dropdowns cannot describe.
3. Size the market before spending a credit
DiscoLike’s count endpoint is free to call with the same filters as a search. PipeRich uses it to size a segment before committing to a campaign, so a segment that is too small to run is caught before any records are pulled.
4. Export, validate, launch
Accounts flow into PipeRich’s own validation step, then into campaigns. Because the count and lookalike steps are exposed through MCP and API, the discovery layer runs inside the same agent workflow that writes and sends the messages.
“It surfaces niche and small-business verticals the usual databases miss, and many of our clients sell into hard-to-find SMB markets.”
Jeff Rolf von den Baumen, Co-Founder & CEO, PipeRich
The Results: Two Months, 56 Clients
- 2.8 million net-new company domains supplied between late July and late September 2026
- About 720 campaigns across 56 clients run with DiscoLike as the primary account source
- Roughly 2x the qualified-account yield of the next-best source (about 1.8x measured), counted as qualified accounts per account exported after PipeRich’s own validation
- Untargeted pockets found in existing client markets through Claude, the Discover MCP, and lookalike search
- Market sizing at zero cost with the free count before each campaign
The yield number is the one PipeRich watches. When several sources feed the same validation step, the source that returns the most qualified accounts per export is the one that saves the team the most time. In PipeRich’s hard verticals, that source is DiscoLike, by about double.
Why DiscoLike Made the Difference
Coverage where LinkedIn-based tools go dark Most of PipeRich’s client markets are SMB-heavy. DiscoLike crawls the business web directly, so a two-person HVAC contractor or a regional insurance agency with no LinkedIn presence is still in the index, with a live, SSL-validated domain.
Search that describes a business, not a category Lookalike search and natural language ICP descriptions let PipeRich define a segment the way a salesperson would say it, then find every company whose website matches. No taxonomy guessing, no custom scripts to stitch together website-level signals.
Built for agents PipeRich’s discovery runs through Claude and the Discover MCP. Count, discover, and validate are the same calls whether a person or an agent makes them, which is what lets one team run 720 campaigns for 56 clients.
Ready to Find the Rest of Your Market?
If your outbound depends on accounts the usual databases miss, DiscoLike can help you:
- Find companies invisible to LinkedIn-based providers
- Size a segment for free before you pull a single record
- Run discovery inside Claude or your own agents through MCP and API
- Turn your best customers into a lookalike list in minutes
Sign up or see pricing.
About PipeRich
PipeRich deploys AI sales workers that generate qualified sales meetings through outbound website contact forms. Its AI identifies and qualifies target accounts, writes a personalized message for each one, submits it through the company’s contact form, and relays interested replies to the client. Founded in Waterloo, Ontario, PipeRich serves B2B teams from seed-stage startups to enterprises. Learn more at piperich.com.
About DiscoLike
DiscoLike is the search engine for the business web. It crawls, indexes, and ranks over 80 million business websites across 55 languages, so people and AI agents can find target accounts by lookalike search, natural language ICP descriptions, and exact phrase matching. Learn more at discolike.com.
Related posts:
