A gut-check before you publish — not a forecast.
Foresight runs a set of synthetic-persona agents against your draft and returns a directional read on likely reach, engagement, sentiment and risk. It is a signal to sharpen the draft with, not a promise of what will happen. We say that plainly here, and again inside the product.
Read this before anything else on the page: Foresight is a directional, synthetic-audience model. It has not been validated against real published outcomes. It is a pre-publish gut-check built to help you improve a draft — it is never a guarantee, a benchmark, or a real forecast of what your post will do.
Persona agents react. Foresight synthesizes.
Foresight instantiates a set of synthetic personas — built from the audience segments you define — and has each one "react" to your draft. It then synthesizes those simulated reactions into one directional read across four dimensions.
- Reach — a directional sense of how far this draft is likely to travel.
- Engagement — whether the personas react, reply or scroll past.
- Sentiment — the emotional tilt of the simulated reactions, positive to negative.
- Risk — flags for tone, sensitivity or claims worth a second look before publish.
A stronger post reads stronger. That is the whole point.
Foresight's read is a function of what is actually on the page — a sharper hook and a clear point of view will out-score a vague, generic version of the same topic. This is a comparison of two illustrative drafts on the same subject, not a real result.
"5 tips for better onboarding."
Generic framing, no specific point of view, a listicle hook that could belong to any company.
| Engagement (directional) | Below baseline |
| Sentiment (directional) | Neutral |
"We killed our onboarding email. Replies tripled."
A specific claim, a clear before/after, a hook that earns the next sentence.
| Engagement (directional) | Above baseline |
| Sentiment (directional) | Mostly positive |
Both examples are illustrative — not results from a real account, and not a claim that either draft actually tripled anything. They exist to show that the read tracks the writing, not the topic.
What Foresight is — and what it deliberately is not.
What it is
- A directional gut-check on a specific draft, before it publishes.
- A synthetic-audience simulation, built from the personas you define.
- A prompt to sharpen the writing before you spend real distribution on it.
What it is not
- A validated forecast benchmarked against real published outcomes.
- A prediction of an actual reach, engagement or sentiment number.
- A substitute for measuring what real audiences actually did after you published.
Foresight is the gut-check before you publish. Analytics is the ground truth after.
Once a piece is live, piMark's learning loop measures what actually happened and feeds it back — that real signal, not the pre-publish simulation, is what teaches the team over time.
See analytics & the learning loopQuestions about Foresight.
Bring a real draft. We'll run Foresight on it, live.
You will see exactly how directional the read is — and exactly what it does and does not claim.