Who is behind it
Because you should ask that about any tool that judges your writing.
smarticulate is built and run by Ian Sharpe, who spent twenty-odd years inside large, complicated organisations turning other people's thinking into something a board would act on. The technical work was rarely the hard part. Getting it said properly was.
Investment banks, law firms, market platforms, in a dozen sectors between New York and Brisbane. The subject changed constantly. The problem did not.
No team behind a curtain. No account manager. Nothing here was bought in and rebadged: the code was written for this, and the method was tested before it was sold.
Where it came from
It was a private tool first
Writing for two businesses with deliberately different voices, and using ChatGPT to help with both, produces a specific and tedious problem. The output drifts. Not badly, and not obviously, but towards the middle. And the middle is where everyone else already is.
So the specification came first, for private use. Then the code, because judgement about writing is easy to fake and a score that changes on a Tuesday is worth nothing. Both were used in anger before either became a product. That order matters.
ConsultingPartner.io
Scores 98 / 100
A fractional executive practice, and the first thing this method was ever run on. Boardroom register, first person, dry, and it argues with the reader on purpose.
Prana Property Care
Scores 93 / 100
Property care and decorating in Cambridgeshire. Chosen because it sits about as far from boardroom prose as writing gets. Plain, short, publishes its prices, never contracts a verb and never once says "I".
The Nail Editor
Scores 95 / 100
An award-winning nail studio in Godmanchester. A different owner, a different writer, and no connection to the two above, which is exactly why it was the test that mattered.
Two of these three were written by the same hand, and their documents came out opposite: opposite pronoun policies, three words apart on average sentence length, contraction rates differing by a factor of seven. Had the method been measuring the writer rather than the business, they would have matched.
The third has no connection to either. It shares none of their structural habits: no semicolons at all, almost no colons, almost no contractions. The rhythm measurements still held, and still rejected the control.
That was the test that mattered. A method which only works on its author's own writing is a party trick.
Three further businesses were measured while building the method, including one chosen deliberately for having weak published copy, plus a generic control written to find out what would slip through. Those three are covered by confidentiality and are not named anywhere, which is the same undertaking every client gets.
Which is why the scoring is published rather than described: how the score works sets out all twenty-two measurements, the four that count, and the ones that were disqualified for ranking the fake above real writing.
What to expect
How the work runs
- Evidence before opinion. If the document says your sentences do something, it shows the count, and the counting code returns the same answer every time it is run.
- The free test is genuinely free and genuinely diagnostic. Sometimes it will tell you that nothing needs fixing.
- Every engagement ends with a document you could use without ever coming back. That is the design, not a concession.
- The aim, every time: Everything your business publishes sounds like your business, whoever wrote it and whatever wrote it.
- Your text is yours. Client work is confidential by default, and nothing becomes a sample or a case study without explicit permission.
The other business
Built by ConsultingPartner.io, run separately from it
ConsultingPartner.io is a fractional executive practice for owner-managed and growing businesses. It built this, and the two are run as separate things. Same person, different question: that one is about what to decide, this one is about how it lands.
Separate clients, separate lists, no funnel between them. Buying from one does not put you in front of the other.