gtmskills.com

Reference file

Fletch execution

fletch-execution.md

Stage 2 — Expression (Fletch PMM)

Source: Anthony Pierri and Rob Kaminski, Fletch PMM, from work with several hundred B2B software companies. See fletchpmm.com and their LinkedIn writing.

Contents

  • The core argument
  • The formula
  • Homepage structure
  • The specificity tests
  • Choosing what to lead with
  • Common copy failures
  • On jobs-to-be-done
  • Working with the user

The core argument

Positioning that lives in a document is not positioning. The website is where positioning is expressed or lost, so the homepage is the deliverable, not a slide.

The most common B2B failure is aspirational abstraction. There is a widespread belief that enterprise buyers prefer high-level, visionary messaging — and plenty of well-known B2B homepages written that way. Read them and you cannot tell what the company does. That works only when brand awareness already answers the question. For everyone else it does more harm than good, because the visitor leaves without knowing what they just looked at.

Specificity is not down-market. Specificity is respect for the reader's time.

The formula

Positioning = target market + differentiation

Expressed as a sentence:

For [SEGMENT] who is trying to [JOB], we're better than [ALTERNATIVE] because [DIFFERENTIATION].

Every element is load-bearing:

  • Segment — specific enough that someone reads it and thinks "that's me" or "that's not me." Both reactions are useful. "Enterprises" is not a segment.
  • Job — what they're trying to accomplish, in their words, from the Stage 0 language inventory.
  • Alternative — what they're doing now. Often a spreadsheet or a manual process, and naming it plainly is more persuasive than naming a vendor.
  • Differentiation — what only you can claim, post table-stakes test.

Drop any element and the sentence stops working. Generic copy is usually a sentence with three of the four missing.

Homepage structure

A workable default. Adapt, don't follow mechanically.

1. Headline — what it is The product category plus the differentiating angle. Not a slogan, not a vision. A reader who knows nothing should finish the headline knowing what kind of thing this is.

2. Subhead — who it's for and how it works The segment and the mechanism. This is where the "how" earns credibility.

3. Social proof strip Logos, a number, or a named result. Placed early because it buys permission to keep reading.

4. The problem, stated in their words Straight from the language inventory. If the visitor doesn't recognize their own situation here, nothing below matters.

5. Three value props, each with a proof mechanism Claim, then how it works, then evidence. A claim without a mechanism reads as marketing; a mechanism without a claim reads as documentation. You need both.

6. Differentiation section Explicitly against the real alternative. Comparison tables work when the dimensions are ones the buyer already cares about, and fail when they're reverse-engineered from your feature list.

7. Objection handling The Stage 0 anxieties and habits: security, migration, "we already have X," "will this replace my team," pricing opacity.

8. CTA Specific and low-commitment. Match the CTA to the anxiety level revealed in Stage 0.

The specificity tests

Run every headline and value prop through all four.

The stranger test. Show the headline and subhead to someone with no context. Ask them to say what the product does and who it's for. If they can't, rewrite. This is the only test that matters and the one most often skipped.

The competitor-swap test. Put a competitor's logo on your copy. Does it still read as true? If yes, the copy describes the category, not you. This kills most "AI-powered platform for the modern enterprise" headlines instantly.

The inversion test. Would anyone credibly claim the opposite? "Secure and reliable" fails — nobody sells insecure and unreliable software. If the inverse is absurd, the claim is empty.

The so-what test. For each line, would a buyer ask "so what?" If yes, you stopped one rung short on the value ladder.

Choosing what to lead with

You cannot lead with everything. Priority order when deciding:

  1. The most specific true thing. Specificity beats scope. A narrow claim that's obviously true outperforms a broad claim nobody can verify.
  2. The differentiator that survived the table-stakes test. If two survived, lead with the one that maps to the strongest push force from Stage 0.
  3. The mechanism, if the claim is hard to believe. When the promise sounds too good, the "how" is the message. "Status without anyone updating it" needs "because it reads your meetings and tickets" immediately behind it.

If the product serves several segments, resist the composite headline that serves none. Either pick the primary segment for the homepage and build segment pages beneath it, or use a headline that names the shared job precisely.

Common copy failures

  • The category-plus-adjective headline — "The intelligent platform for modern operations"
  • Naming a coined internal term in the headline — the visitor has to learn your vocabulary before understanding the product, and they won't
  • Value props that are feature names — "Smart Dashboards" tells the reader nothing
  • Three value props that are the same value prop — a sign the value ladder wasn't done
  • Proof that proves adoption rather than the claim — user counts prove habit, not that the product does what you say
  • Hiding what it is below the fold — if "what is this" is answered in section four, the headline failed
  • Personas instead of segments — "For innovative leaders" is flattery, not targeting
  • Copy written for the board — investor language leaking into buyer-facing pages is extremely common after a fundraise

On jobs-to-be-done

Fletch deliberately keeps JTBD vocabulary out of their own framework, because the community is split between two incompatible definitions of a "job" and the term generates more debate than clarity.

This skill uses JTBD in Stage 0 as a research method and Fletch's formula in Stage 2 as an expression method. That is a compatible pairing, but do not present Fletch as a JTBD framework. If a user asks, explain the distinction rather than blending them.

Working with the user

Expect resistance at this stage. Specific copy feels risky — it forecloses segments, it makes claims that can be checked, and it sounds less impressive read aloud in a leadership meeting. The abstraction almost always comes back in the review cycle.

Two things help. First, show the competitor-swap test live: put their draft next to a competitor's logo. Second, remind them the homepage's job is to help the right visitor self-identify and the wrong visitor leave. Broad copy achieves neither.

Write two or three headline options at different specificity levels rather than one, and let the user see the tradeoff rather than argue about it in the abstract.