Amplification ladders

Read this when a gate criterion scores "no". Each ladder runs weakest to strongest. Climb until the criterion bites, then re-score.

Risk (criterion 3)

The buyer should not be able to lose. Climb until that's literally true:

1. A bare claim — "we get results"
2. Money back if unhappy
3. Pay only on results
4. Work free until the first win lands
5. The value is delivered free, in full, before any commitment

Rungs 4 and 5 are the only ones that reliably work when you have no brand. If guarantees are contractually banned, rung 5 is the answer: the free asset must be so complete that judging it costs the prospect nothing.

Proof (criterion 6)

1. No proof
2. A vague claim — "we get great results for teams like yours"
3. A hard number, anonymized — "a lender doing $100k/mo: 5–10 booked demos a month"
4. A hard number attached to a named company
5. A named company that visibly resembles the prospect, plus a number, plus a way to verify it

The closer the comparable resembles the reader, the more it converts. Put it inside the offer. "Happy to send a case study" defers the single element that earns belief, and the reply that would have asked for it never comes.

If naming clients is contractually banned, rung 3 is your ceiling — still concrete, still a real number, never "we get great results."

Effort (criterion 4)

1. They have to learn something, or wait
2. Do-it-yourself with your guidance
3. Done with you
4. Done for you
5. Already done, sitting there, one reply away

A meeting is a cost you charge the prospect, not a gift you give them. Pre-build the value from public data or from data they can send asynchronously, deliver it, and earn the meeting afterward.

Specificity (criteria 1, 2, 6)

Vague reads as unbelievable on cold traffic, because a stranger has no context to fill the gaps with goodwill.

• "more replies" → "32% more replies in 47 days"
• "faster ramp" → "new reps productive in 5 weeks instead of 11"
• Odd numbers and stated timeframes read as measured; round numbers read as marketing.
• Every claim should be traceable to something you could show if challenged.

Mechanism (criterion 2)

Name it. Tie it to something others structurally can't copy — a data source you built, a process only you run, an insight from work only you have done. Then answer the skeptic's two questions plainly: why does this work and why hasn't it worked for me before?

A mechanism that could be claimed verbatim by three competitors isn't a mechanism.