Skip to content
RAY BEHARRY

Agent 8 of 71 · Positioning & Strategy

Differentiation Analyst

Separate real difference from parity. Map what each competitor claims, what buyers say they care about, and what we can prove, then surface the white space no competitor occupies that buyers value and that our evidence can defend.

Agent contract

Use to find what we can credibly claim that competitors do not: narrative white space, parity claims to stop leading with, and proof-backed differentiators. Produces a differentiation map.

When it runs

  • A positioning refresh needs a validated Differentiator variable.
  • Competitor Monitoring Agent logs a messaging change that may close an existing gap.
  • Win/loss notes show we lose on 'they all sound the same'.

What it needs

InputRequiredFrom
Competitor claims and messaging historyYesCompetitive Intelligence
Buyer pains and verbatim phrasesYesMarket Signals
Proof records we can citeYesProof Vault
Current Differentiator in canonOptionalPositioning Canon

How it works

  1. Build a claim grid: rows are buyer-valued themes from VoC, columns are us and up to five competitors; each cell holds the public claim text and its URL.
  2. Mark each theme as table stakes (three or more players claim it), contested (two claim it) or open (no one claims it).
  3. Score buyer importance per theme using VoC frequency counts, not internal opinion.
  4. For open and contested themes, check whether a ProofID supports our claim; a theme without proof is white space we cannot yet own.
  5. List parity claims we currently lead with and recommend demoting them below the fold.
  6. Write the differentiation-map with three ranked white-space candidates, each with buyer importance, proof status and defensibility window.
  7. Log the map date so the next run can show which gaps competitors closed.

What it delivers

Claim grid (theme × competitor with sourced claims), theme status, buyer-importance counts, three white-space candidates with proof status, parity claims to demote.

Output contract: differentiation-map

Quality bar

  • Every competitor claim is quoted from a public page with URL and capture date.
  • Buyer importance comes from counted VoC evidence; themes without counts are hypothesis.
  • A white-space candidate without an unexpired ProofID is labelled 'needs proof', not 'differentiator'.

Hands off

When a person steps in

  • If our strongest differentiator depends on a roadmap item, flag it and escalate; never present unreleased capability as a differentiator.
  • Escalate when a competitor's claim appears to misrepresent our product, for legal review rather than public rebuttal.

Example

Request

“Where can we stand apart from the four AP vendors we lose to?”

Result

Illustrative output: 11 themes mapped; 6 table stakes (OCR accuracy, approvals). 'Audit trail a regulator accepts' is open, ranked 2nd by VoC (19 of 60 docs), backed by PRF-0022. 'Fastest setup' is contested and lacks proof, flagged 'needs proof'. Recommended demoting 'AI-powered' (claimed by all 5).

Illustrative example from the package documentation, not a client result.