Let me be Brief...

The research nobody reads twice

Before a first call, I do what most growth people do. I open the company's site. I read the pricing page they read. I guess at the tools behind the page from the source. I find the two or three people who will be on the call, or should be, and I write a paragraph on what I would say to them. If I am disciplined it takes an hour. Then the call happens, and the hour is gone. Nobody reads that research twice, including me.

Bryn has been doing that read for us the whole time, on every visitor to our site, in the time a page takes to load. Nobody had asked it to write the read down.

Let me be brief. Now it does.

What a Brief is

Bryn's job is four steps. It watches your site and product for a signal: something a buyer just did, like coming back to your pricing page. It scores whether the account behind the signal is worth acting on. It runs the follow-up you approved, under a rule you set once: when this happens, do that. Bryn calls that a Play. And it logs every run so whoever signs off can read it. Watch, score, act, log.

A Brief is that read on one company, on request, written out. The sign-in page says it in one line: "Bryn scores the buyers on your site and runs the follow-up; a Brief is that read on one account."

Name a company, yours or one you are about to talk to. Inside a minute Bryn hands you a page about it, with a copy by email. You can pass it to whoever signs off. The hour I described above never produced that.

What is in one

Six sections, set in the company's own colours. Here is what each one is for.

Who it is, and whether it looks like a buyer. What the company sells, how it sells it, how big it is and where it is, read from its site. Then the fit read: does it look like the customer Bryn is built for.

The score out of 100. What Bryn looked at (industry, headcount, how the company sells, geography), what it found, and one line on what would move the number. When the match is thin it prints "Partial fit" and what is missing, in place of a number.

The score is Bryn's read of whether this company fits the customer Bryn is built for. In this preview, the customer is Bryn: every Brief scores a company against the customer Bryn is built for. The same read, scored against your customer, is being built into Bryn. The preview is out now so you can watch it work first. Run it on your own company and you will see the read Bryn gives every visitor to its site.

The tool stack, ranked by what it means for the follow-up. Every tool found on the site, sorted into what Bryn connects to live today, what it would replace, what rides on the tag manager, what a webhook reaches, and what is on the roadmap. Each row carries a word on where it was proved: on the page, in the company's own domain records, or in its copy. The integrations page puts the offer this way: "A Brief reads your site and names what's already running, what connects today, and what Bryn would replace."

The people who would sign off. One card each, with title and LinkedIn.

The first follow-up Bryn would run. The Play Bryn would run first for this company, matched to how it sells. For a company Bryn already watches, what runs next, each Play with the signal that sets it off, the ones its traffic has already tripped first.

"This page is a Play that already ran." The receipt. More on it below, because it is the section I would read first.

Brynby CivicFigure 1 · one Brief, section by section

One Brief, six sections.

Pick a section. The sheet shows that part of the Brief for an invented company, and the line under it says what you do with it.

1. Who it is, and whether it looks like a buyer. What the company sells, how it sells it, how big it is and where it is, read from its site (illustrative rows), then the fit read as a meter with the gloss looks like a buyer (illustrative). What you do with it: decide in one glance whether this company is worth the call.
2. The score out of 100. A meter out of 100 (illustrative), the reasons as rows (industry, headcount, how the company sells, geography) with what Bryn found and the effect, and one line on what would move the number. When the match is thin the Brief prints "Partial fit" and what is missing instead. What you do with it: disagree with a reason, not with a number.
3. The tool stack, ranked by what it means for the follow-up. Rows for what connects live today, what Bryn would replace, what rides on the tag manager, what a webhook reaches, and what is on the roadmap, each with a word on where it was proved. What you do with it: know what the follow-up has to fit into before you write it.
4. The people who would sign off. One card each, title and LinkedIn, initials only here. What you do with it: write to the person who owns the number, not the inbox that answered.
5. The first follow-up Bryn would run. The first Play, matched to how the company sells, then what runs next with the signal that sets each off (illustrative). What you do with it: run it as written, or edit the one line you disagree with.
6. This page is a Play that already ran. The rule in four lines (request watched, company scored, page and email as the action, record written), the trace with clock stamps and the elapsed time, and one receipt row per section as it landed (illustrative clock times). What you do with it: read it before the score. Every row is something you used to do by hand.

Illustrative company (quillfeather.example), people shown as initials and titles. Every value and every clock time is illustrative.

The Brief is itself a Bryn run

Briefs is the first agent built on Bryn: the same four steps Bryn runs for customers, watch, score, act, log, pointed at Bryn's own front door. In plain words: Briefs is Bryn pointed at its own website, started by your request instead of a visitor's.

Your request is a signal, and Bryn watches for it. Bryn reads the company's site and scores it, with the same weights and the same bar a customer's own account would have to clear. It builds the page, sends the email and posts its own follow-up to Slack. Then it writes the record. The last section of every Brief says so, under the heading "This page is a Play that already ran", and lays out the rule that built it in four lines: the request was watched, the company was scored, the page and the email were the action, the record was written.

Under those four lines runs the clock. The Play trace stamps each step from your request to Bryn's follow-up in Slack, with the elapsed time. Under that sits the receipt: one row per section, stamped as it landed. The page is its own proof.

You don't have to check Bryn. Bryn checks in with you, once the work is done.

That receipt has two readers. The first is you, before the call. Every row on it is a piece of that hour, done while you watched. The second reader is whoever signs off. We have written before about rules you can read in plain words, in Your Rules, Readable, and about a record another person can reconstruct a run from, in Show Me a Blocked Action. A Brief keeps the same kind of record about itself. Nothing in it is a conclusion without a step behind it, and every step has a time on it. Yesterday's note made the case that the decision, and the record of it, is the layer you cannot rent; a receipt at the end of a document is a small example of what that looks like. Later this week we walk the receipt line by line.

The score keeps moving after the send

The request page asks you to start with your own company. That Brief is the one that closes a loop.

Open a Brief a week after it was sent and the score is re-read from Bryn the moment the page opens. The Brief keeps the score it was written with beside it, so you can see the distance. Under the score sits a list headed "Since this Brief was sent": what the company's people did since, in plain lines like "Someone from your team visited pricing". Each of those lines is a signal, and it went into Bryn the same way every other one does.

One rule about whose record it lands on. When someone at company A asks for a Brief on company B, the interest belongs to A. The request is filed on A's record, and it is A's own Slack rule that fires. B's record stays B's own.

So the read is not finished when the document is. You asked, Bryn wrote, the page went out, and what happened next came back as evidence on the page itself.

Brynby CivicFigure 2 · the score keeps moving

One Brief, from the request to the re-read.

Step through the moments. The account card shows which of three plain reads moves, and the log line that records it.

1. The request lands on the asker's record (09:14:02, illustrative). Someone at Quillfeather Labs asks for a Brief. The request is a signal, filed on Quillfeather's own record; had they asked about another company, it would still be filed on Quillfeather, and Quillfeather's own Slack rule would fire. Before the Brief: looks like a buyer, not yet acting like one. Log: brief requested · filed on quillfeather.example
2. The Brief is sent (09:14:49). The page and the email go out, inside the minute. The Brief carries the score it was written with. Nothing has moved yet. Log: page and email sent · score as written
3. The page is opened later and the score re-reads (10:02:11, a week later). The score is re-read from Bryn the moment the page opens; the written score stays beside it. Acting like a buyer moves; acting now moves. Log: page opened · score re-read from Bryn
4. "Since this Brief was sent" lines appear (10:02:12). Under the score: Someone from your team visited pricing. Someone from your team came back to the docs. (illustrative) Each line is a signal that went into Bryn the usual way. Log: since this Brief was sent · listed under the score

Illustrative account and values. Clock times are invented. The Brief keeps the score it was written with; the re-read happens the moment the page opens.

Public Preview: how to get one

Go to civic.com/briefs and sign in. Any email address signs in. Name a company. The request page introduces itself first, in the only line in this piece where Bryn speaks for itself: "I'm Bryn, the GTM Agent." GTM is the sales and marketing side of a company. Then, in its words, "Your Brief builds while you watch." The sections land one by one, inside a minute, and a copy arrives by email.

Preview means what it says. Briefs is open, it is being worked on in the open, and it may change week to week. If a section moves or a step finds a better source between one week and the next, that is the preview doing its job. The receipt will tell you what changed, because the receipt is the record of how the Brief was built.

Then run one on a company you are about to talk to. Read the receipt before the score. Every row on it is something you used to do by hand.

Why we built it this way

Civic is Bryn's first customer. Before Bryn ran for anyone else it ran for us, and it still does; I told that story in We Are Customer Zero. Briefs is the agent we built for ourselves on top of it, on Bryn's API and its MCP server, the same connection your own agent can plug into. It is the research we used to do by hand before a first call, done the way Bryn already does it for every visitor, and handed back as a page that proves how it was made. Last week's note on the rebuilt app made the case that a score should always show its parts. A Brief is that case in a form you can forward.

Bryn is not another dashboard to watch. It is the governed execution layer that runs Plays through your stack. In plain terms: you write the rule once, Bryn carries it out inside the tools you already use, and nothing runs without a record of who approved it. A Brief is the fastest way I know to see what that layer sees, on one company, without installing anything.

Chris Hart, our CEO, is at Clay's GTM conference in San Francisco this week, on Thursday. If you meet us there, the card in your hand has a code on it; scan it and you land on Briefs.

Stop watching signals. Start running them.