ABM at scale: running 1,000 accounts like they're 30

The phrase 'ABM at scale' used to be a contradiction — the M stood for manual. What changed isn't the strategy; it's that the most labor-intensive layer (per-account experiences) became generative. Here's the operating model for covering a four-figure list with a team of two to five.

Layer 1 — Segmentation that maps to templates

Split the 1,000 by the problem your champion feels, not just industry. Each segment gets one template whose fixed copy (problem framing, proof, offer) rings true for every account inside it, and whose slots (company name, researched situation, industry stat, brand colors) the engine fills per account. If a template's pain section wouldn't survive being read by all 200 accounts in the segment, the segment is too wide.

Layer 2 — Generation and QA at volume

• Batch-generate by segment; spot-check 10% — headlines and research lines, since structure is template-fixed.
• Set confidence gates: pages where research came back thin get flagged for human review, not silently shipped.
• Regenerate on trigger events (funding, leadership change, product launch) so pages stay current without manual sweeps.

Layer 3 — Distribution through what you already run

1. Sequences: the page replaces the homepage link in every step.
2. Paid: LinkedIn ads to the account, landing on its own page — ad-to-page message match no campaign page achieves.
3. Direct mail and events: QR codes resolving to the account's page.
4. Sales touches: AEs send sections, not attachments — 'the integration part is the second block on your page.'

Layer 4 — The weekly operating rhythm

1. Monday: engagement review — accounts sorted by page intent (returns, multi-viewer, CTA clicks). Promote the top decile to human plays.
2. Tuesday–Thursday: humans run promoted-account plays; the engine handles everyone else's follow-ups.
3. Friday: template health — which segment's pages convert visits to meetings, which need a new proof block or sharper pain framing.

The headcount math is the point: the old model spent team hours proportional to accounts covered. This model spends hours proportional to accounts engaged — a number an order of magnitude smaller and infinitely better qualified.