$ man clay-wiki/contact-cards

Contact cards are enriched contact-level profiles built in Clay for persona qualification and outreach routing. Where company cards answer "is this company a fit?", contact cards answer "is this person the right one to talk to?" You source contacts from qualified accounts, enrich each one with title, department, LinkedIn data, and persona tier, then route them to the right platform for outreach.

PATTERN

Contact cards always sit on a separate table from company cards. The contact table sources people from qualified accounts on the company table. Every contact row has a lookup column pointing back to the account table — pulling in ICP score, service fit, industry, and any company-level research. This is the account-contact split pattern. Never enrich both levels on the same table. The flow: Company cards → qualify accounts → source contacts from qualified accounts → contact cards → qualify personas → route to outreach.

PATTERN

Every contact card should have these columns: • Email (from single-provider lookup — Prospeo or LeadMagic) • First name / Last name • Job title • Department • Seniority level • LinkedIn profile URL • Persona tier (1-5 from qualification prompt) • Outreach priority (primary / secondary / champion / not_target) • Company domain (lookup key to account table) • ICP score (pulled from account table via lookup) • MX provider (Google / Microsoft / Other — from HTTP column) • Route (Instantly / HeyReach / skip) • Personalization variables (icebreaker, pain point — from research prompt)

PATTERN

The persona qualification prompt maps job titles to buyer tiers. Every partner has 3-7 persona tiers defined: • Tier 1 — Economic Buyer (C-Suite, SVP). Primary outreach priority. • Tier 2 — Technical Decision Maker (VP, Director). Primary outreach priority. • Tier 3 — Influencer / Champion (Manager, Senior IC). Secondary priority. • Tier 4 — Operational (Specialist, Analyst). Low priority or skip. • Tier 5 — Edge cases (varies by partner). The prompt takes a job title and outputs: persona_qualified (MATCHED / NOT_MATCHED), tier, tier_name, outreach_priority, and reasoning. A clean persona qualification prompt eliminates 40-60% of contacts before you spend credits on email lookup and personalization research.

PRO TIP

After persona qualification, run the email lookup only on MATCHED personas. Use one provider — Prospeo or LeadMagic. Don't stack a 6-provider waterfall. I used to run Apollo → Hunter → Clearbit → RocketReach → Prospeo → Dropcontact. I stopped because the routing logic made the waterfall pointless. Here's why: once you check MX records and route Google → Instantly, non-Google → HeyReach, the marginal coverage from extra providers doesn't matter. You're burning credits chasing emails that bounce anyway when the MX doesn't match your sending infrastructure. One provider + MX check = same deliverability at a fraction of the cost. MX records tell you more about deliverability than any validation tool.

ANTI-PATTERN

Running persona qualification after the email lookup. Flip the order. Qualify the persona first. If the title doesn't match any tier, don't waste credits finding their email. You'll never contact them anyway. Another mistake: sourcing contacts without checking the CRM first. Before any contact enrichment, run a CRM lookup. If this person is already in HubSpot with an active deal, don't re-enrich and don't add them to a cold outreach sequence. Sending leads already in pipeline is the fastest way to lose trust.