What Apero builds in Clay

What we build in Clay

Clay is list and research. We use it to build the account universe, score it, enrich it, and route it to the channel that fits.

Ben Holley's take on why it stays in the stack: it is deterministic, and you can audit every row. When a team needs to see the work, the work lives in Clay.

Apero runs Clay's CLI to execute stock workflows deterministically (company research, buying committee mapping, email finding), wired into Grok Bot and Claude Code, so agents call a proven workflow instead of reinventing the research from first principles each time.

Clay vs Vercel rule: Apero uses Clay for email and LinkedIn lookups, triggered with webhooks or HTTP API requests, and builds on Vercel for flexible, non-traditional sales workflows. Example: Apero's network mining tool is built outside Clay.

Example workflows

1. 01

   Build one deduplicated TAM

   Trigger

   A company needs every account it could realistically sell into, not just what a database filter returns.

   System

   Lists from Clay, Serper, and company websites are deduped into one universe. Firmographics are added, then each account is qualified from its website and job posts.

   Output

   One account universe with the data points that decide fit, instead of a pile of competing spreadsheets.

   Operator outcome

   Every channel downstream works from the same list, so channel results can actually be compared.

2. 02

   Score accounts against many use cases

   Trigger

   A product has dozens of valid buyer types and one vertical tag does not describe fit.

   System

   Fit criteria are written per use case with the client. Research agents (often Clay plus Exa or custom scrapers) read websites and integration pages and pattern-match against known customers.

   Output

   For each company: what they do, how they could use the product, and which product lines they should buy, with the reasoning visible.

   Operator outcome

   At Trackstar, sales hires prospect from scored lists instead of founders opening exhibitor websites by hand.

3. 03

   Turn a conference list into targets

   Trigger

   An attendee or exhibitor list arrives with first name, last name, company, and title.

   System

   The list runs through the same scoring as cold outbound. Each person gets fit, email, LinkedIn, CRM overlap, and mutual connections.

   Output

   A short list of who is worth a meeting and why, before the show.

   Operator outcome

   Limited LinkedIn slots and meeting time go to the highest-confidence fits.

4. 04

   Score the CRM you already have

   Trigger

   A Salesforce or HubSpot instance is full of accounts nobody has looked at in a while.

   System

   The CRM is pulled into a Clay table, an agent reads each website, and accounts are scored 1 to 4 with a short note an AE would actually read.

   Output

   A tiered account list with a written reason per account.

   Operator outcome

   Reps stop treating the CRM as a graveyard and start from accounts with a reason to be there.

5. 05

   Enrich, gate, and route to the right channel

   Trigger

   A segment is ready to go out by email, LinkedIn, or phone.

   System

   Enrichment waterfalls add firmographics, hiring, and tech stack. Table logic routes each account: the right SDR sequence, skip email where a security gateway blocks cold mail, LinkedIn-only when that is the only path.

   Output

   Segments pushed to EmailBison, Lemlist, or HeyReach, and dial lists that refresh as titles change.

   Operator outcome

   Fewer sends into filters, and lists that do not go stale every quarter.

How we work in your Clay workspace

Access

Client's choice: Apero builds in the client's Clay workspace, or in Apero's. Either way Apero needs access to the CRM and senders the motion needs.

Change control

Shared work that the team needs to see, like a TAM refresh, stays in Clay so anyone can audit a row. Logic that has to be durable or connect several systems moves into a repo with reviewable code.

Documentation

Fit rules and ICP definitions are written down next to the scoring, so the next person can see why an account was kept or skipped.

Handoff

Operators are trained to rerun and refresh the tables. Data degrades fast, so we plan a refresh cadence (ideally quarterly) as part of the handoff.

When we do not use Clay

• Heavy one-off personalization. Benyamin (Ben) Holley, Apero's Head of GTM, finds Clay slow for it because you read every cell.
• Anything that needs persistent business context across steps. Clay does not keep it, so that work lives in Cursor or Claude Code.
• Durable systems like reply agents, CRM health checks, and integrations. Those are software.

FAQ