Finds the companies worth contacting now. Then keeps finding them.
Right company, right moment, right message. A flow that grows every time you run it.
Orbit is the part of Studio that finds companies where the problem you solve has just become live, stages the message in your voice, and keeps finding more like the ones that reply.
By the time it says go, it has been watching for weeks.
The event that tips it is the last thing it learns, not the only thing it knows.
One job posting on its own means very little. The same posting, read against four months of quiet observation and against the specific problem you solve, is the difference between a guess and a reason.
Fourth signal pointing the same way, and the first one with a budget behind it. The problem you solve just became someone’s job there.
Every company it has met, and where each one stands.
Fired, on the watchlist with its gap named, or discarded with a count. It tells you why for every one.
Nothing hidden, nothing on a hunch. A near miss is one reason away: say the word and it drafts. No company is researched twice.
It works out what a buying moment looks like from what you sell.
No keyword lists. No alerts to maintain. It watches for the marks your problem leaves in the world.
The problem is about to exist.
Funding. A key hire. A new market. A regulation taking effect.
The problem exists now.
A job spec naming the bottleneck. A public complaint. A hiring pattern.
It is top of mind today.
Someone says it out loud, in a talk, a post or an interview.
Each one carries where it can be seen.
A search, a careers page, a CRM field. Without that, a footprint is a wish rather than an instruction. Change what you sell and the whole profile rebuilds.
Finding a new company is the expensive part. So it works in batches you can stop.
Five in, a verdict on each, then next batch? Never a list without a verdict.
Funding round-ups, hiring pages, launches: the places your problem leaves a mark. Every batch reports what fired, what was logged, what was discarded.
A win becomes a seed. Find more like Estel, and it keeps producing while the vein holds. No daily cap.
Never researched twice.
One index, checked before any search is spent.
Routes to that contact’s cooldown. Never a fresh cold message to someone you have already written to.
A quick check of whether the gap has closed. The dossier already exists, so the question is only what changed.
Skipped at no cost. Re-opened only if the company has crossed into your band since.
It gets smarter as replies come in. And it will not learn from six of them.
Two loops that can see each other. A thin streak is a watch, never a rule.
What you fire and dismiss tunes what it surfaces. A pattern becomes a question to you, never a rule it sets itself.
Replies and outcomes, read against your targets. Silence stays unknown. A win names its next move: find more like them.
Every move lands ready to send. Volume is a reading, not a brake.
A named person, an address that resolves, waiting in the tool you already send from.
The failure this exists to prevent:
a finished draft addressed to a job title, with the real person still to find. That is homework, not a move. Approving should take one action, not open a list.
No daily cap. Staged this week, sent this week, in front of you. The bar does not move: volume comes from finding more companies, never from clearing them more easily.
Nine checks clear before anything reaches you.
Seven ask whether there is a move here. Two ask whether the words are safe to send.
It dates when the thing happened, not when it found it.
Five known ways a page hands you a wrong date. It checks all five, and says so when a date cannot be established.
It keeps how loud and how sure apart.
Merged into one score, a loud signal nobody understands looks the same as a quiet one you do. They are not, and they earn different tone.
Every claim traces to approved proof, or it is cut.
A logo you may only use in a case study is never cited cold, even when the company reads exactly like that customer. That resemblance is the temptation the rule removes.
It never writes down a guess about what someone privately wants.
Not even as a hypothesis. Noted once as a maybe, a guess reads as background fact six weeks later — and by then nothing tells it apart from something observed.