The unit economics of personalized pages: cents per page, dollars per meeting

Every personalization debate eventually hits the same wall: 'we can't afford that for every account.' Fair: under the old cost structure. The structure changed, and most teams' mental math hasn't. Here are the actual numbers, by production method.

Method 1: the hand-built page

Research (30–60 min), copy (30–60 min), design and build (30–120 min): call it 1.5 to 4 hours across rep, marketer, and designer. At loaded costs, $150–$600 per page. Quality ceiling: highest. Throughput ceiling: a handful per week, which is why this method built ABM's tier system in the first place.

Method 2: the rep-DIY template clone

Duplicate a deck-style page, swap the logo and a paragraph: 15–30 minutes of rep time, $15–$40 loaded. Quality: visibly thin (prospects pattern-match clones as fast as they pattern-match merge tags). The economics aren't actually good; they're just visible on no budget line.

Method 3: the generation engine

• Template: built once per segment (hours, amortized to ~zero per page).
• Per page: automated research, slot-fill, brand extraction, render, marginal cost in cents.
• Concrete: $99/mo ÷ 600 pages ≈ 16.5¢; $299 ÷ 2,000 ≈ 15¢; $799 ÷ 6,000 ≈ 13¢.
• Human time returns only for QA sampling and intent follow-up, the parts humans are actually for.

Modeling it against revenue

Worked example, deliberately conservative: 1,000 prospects/month at ~15¢ = $150. If pages lift email→page→meeting conversion to just 1% of list (10 meetings), cost per sourced meeting is $15: against the $200–$700 blended cost most outbound programs carry. Even at a quarter of that lift, the program clears its bar. The sensitivity that matters isn't page cost (it's rounding error); it's whether your list and template deserve each other, which is where the effort actually belongs now.

When personalization cost dollars, the question was which accounts deserve it. At cents, the question inverts: which accounts can you justify leaving generic?