$ man how-to/how-to-build-abm-pipeline-with-ai

Traditional ABM requires Demandbase ($50K+/year), a dedicated ABM manager, a content team, and months of setup. That is enterprise ABM. Most startups cannot afford it and do not need it. AI-powered ABM flips the model. Instead of expensive intent platforms and manual account research, you use Clay for enrichment, AI for personalization, and automation for execution. A solo operator can run a targeted ABM program that outperforms enterprise teams by being more precise and faster to iterate. The pipeline I built generates qualified meetings from target accounts at roughly $525/month in tooling costs. Here is the complete system.

PATTERN

Most teams define their ICP in a slide deck: "Mid-market SaaS companies, 50-500 employees, Series A-C, North America." That is too vague to automate. An automatable ICP is a scoring formula. Every attribute gets a point value: Company size: 50-200 employees = 10 points, 200-500 = 8 points, 500-1000 = 5 points. Funding stage: Series A = 10, Series B = 8, Seed = 5, Bootstrapped = 3. Industry: SaaS = 10, Fintech = 8, E-commerce = 5. Tech stack signals: uses Salesforce = 5, uses HubSpot = 3, uses Outreach = 5. Hiring signals: posted SDR/BDR jobs in last 90 days = 10 points. Growth signals: headcount grew 20%+ in last 6 months = 8 points. Total score determines tier: 40+ = Tier 1 (priority outreach), 25-39 = Tier 2 (standard outreach), below 25 = skip. This formula goes directly into Clay. Every account gets scored automatically. No human judgment needed for the initial sort.

CODE

The enrichment pipeline runs in Clay with waterfall logic: 1. Source accounts from Apollo by ICP filters (industry, size, location, tech stack) 2. Enrich company data: revenue (from Apollo), tech stack (from BuiltWith via Clay), funding (from Crunchbase), headcount growth (from LinkedIn via Clay) 3. Score each account using the ICP formula 4. For Tier 1 and Tier 2 accounts, find contacts: pull decision-makers by title (VP Sales, Head of Revenue, CRO, VP Marketing) 5. Waterfall email enrichment: try Apollo email, then Prospeo, then Dropcontact until verified 6. Validate emails: MX record check to confirm the domain accepts email 7. AI research: use Claygent to visit each company's website and extract one relevant personalization datapoint (recent product launch, hiring surge, funding round, partnership announcement) The entire pipeline runs in one Clay table. New accounts flow in, enrichment happens automatically, and qualified leads with verified emails and personalization data flow out to your outreach tools. Cost per enriched lead: roughly $0.50-1.50 in Clay credits depending on how many enrichment steps fire. For 500 leads/month, budget $250-750 in Clay credits.

PATTERN

Qualified leads from Clay get pushed to two outreach channels simultaneously: Email via Instantly: 4-step sequence over 14 days. Email 1: personalized cold email referencing the Claygent research. Email 2 (day 3): value-add follow-up with relevant insight. Email 3 (day 7): social proof or case study reference. Email 4 (day 14): breakup email. LinkedIn via HeyReach: parallel 3-step sequence. Day 0: profile view. Day 1: connection request with personalized note. Day 3 (if accepted): direct message expanding on the email. The timing is deliberate. Email hits first, LinkedIn reinforces. The prospect sees your name in two places within the same week. This multi-channel approach typically generates 2-3x the response rate of email alone. Personalization at scale: the Claygent research from Step 2 feeds into both email and LinkedIn templates as merge variables. Each prospect gets a unique first line that references something specific to their company. This is not "Hi {first_name}, I noticed {company_name} is growing" - it is "Saw your team just launched the enterprise tier and posted 3 SDR roles this month. That scaling pattern is exactly where our automation fits." Response handling: positive replies go to the CRM (Attio) as new deals. Meetings booked via Calendly link in the email. Negative replies trigger automatic removal from all sequences.

PRO TIP

The complete ABM stack: Clay ($349/mo) - enrichment, scoring, personalization research Apollo ($49/mo) - contact database and initial sourcing Instantly ($47/mo) - email delivery and domain rotation HeyReach ($79/mo) - LinkedIn automation and account rotation Attio (free) - CRM and pipeline tracking Google Workspace ($6/mo per mailbox x 6 = $36/mo) - sending infrastructure Total: roughly $560/month What this produces: 500-1000 enriched, scored, personalized outbound contacts per month. Multi-channel sequences running automatically. 24/7 pipeline generation with minimal daily maintenance (30 minutes for reply handling and sequence optimization). Compare to: one SDR costs $70,000-90,000/year ($6,000-7,500/month). This automated pipeline produces comparable or higher meeting volume at 8% of the cost. And it scales with tooling costs, not headcount. The human element: you still need someone to handle positive replies, run discovery calls, and close deals. The automation handles everything before the first human conversation. That is the 80% of the sales process that used to require manual SDR work.