I treat language like infrastructure. Every public claim is load-bearing.
If a claim can’t hold under scrutiny, it becomes a liability. If an investor or partner can’t repeat it accurately, it can’t travel.
This page sets out how I work, and the claims integrity standard I apply to every engagement.
How I work (clean, bounded, repeatable)
I keep the process light on meetings and heavy on concrete outputs. One deliverable at a time. One revision round.
-
1) Intake
What happens : I take your current asset (page, deck, doc) and the context it sits in : audience, stage, and desired outcome.
Output : Inputs confirmed + definition of done (in writing).
-
2) Audit or Draft
What happens : I'll either, tag or remove jargon that hides the mechanism, and provide language that overstates it clearly.
Output : Defined Asset of drafted deliverable in a Google Doc with clear section structure.
-
3) Claims integrity check
What happens : I tag key claims, add boundaries, and note what evidence supports each claim. If evidence is missing, I qualify or remove the claim.
Output : A short claims ledger (claim : evidence : boundary : status).
-
4) Review and revisions
What happens : We agree on scope of revisions for each project. You review the draft and apply your notes. I apply the revisions. If necessary we repeat, and if new scope appears, I’ll flag it before doing extra work.
Output : Final copy + access to revision history (what changed, what was clarified).
-
5) Handover
What happens : I leave you with the necessary documentation you can work from or implement.
Output : Final doc + accompanying documentation.
Claims Integrity (Defensible by Design)
- I separate language by confidence level : verified / evidenced / modelled / aspirational.
- I write with boundaries : scope, conditions, time-bounds, and measurement method where relevant.
- I ask to see supporting material for key claims before they go public.
- I qualify what can’t be proven yet. If it can’t be qualified cleanly, I remove it.
- I do not use vague virtue words as a substitute for mechanism or evidence.
Evidence levels (how I label reality)
When a claim matters, I tag it. This prevents accidental overclaim and makes review fast.
I’m not here to make you sound smaller. I’m here to make you sound trustworthy.
What I need from you (to keep claims defensible)
This keeps the work truthful without turning it into a research project.
- Your preferred terms and any terms you avoid (internal lexicon, even if rough).
- Links to the sources behind key claims (reports, datasets, methodology notes, monitoring results, partner documentation).
- Your current boundary definitions (what’s included, excluded, and where uncertainty sits).
- Any claims you consider sensitive (regulatory, reputational, procurement, investor diligence).
- I will provide an onboarding form or document before we begin.
What I won’t do
- I won’t invent metrics, imply verification that doesn’t exist, or write broad ‘green’ claims without boundaries.
- I won’t provide legal advice or guarantee regulatory compliance. You own final sign-off on factual accuracy.
- I won’t trade clarity for cleverness or pressure tactics.
What you get
- Language that holds up in investor rooms, partner conversations, and diligence.
- A story that can be repeated without distortion.
- Copy that stays maintainable as your data and scope evolve.
A small example (weak claim : strong claim)
“We provide sustainable solutions for water and soil.”
“We integrate on-site renewable energy with biochar systems to reduce diesel use and improve soil carbon conditions, measured against site baselines (as of [date]).”
Same intent. The second version names the mechanism and adds boundaries. It gives the reader something they can interrogate.
Start with a bounded micro-sprint
If you want to test fit, I start with one small, high-leverage deliverable. Fixed scope. One revision round. Timeboxed delivery.
…Conscious Copy