How it’s generated
What goes into an edition
A briefing is built from thirteen dimensions, and they arrive by two different routes. The five monitored dimensions have been running on their own schedules since you set the project up, so the briefing reads their accumulated history — weeks of checks it can look back across. The other eight aren’t polled on a timer; CompetLab researches them fresh each time it compiles an edition, because they’re the slower, higher-context signals that read better as a periodic study than a live feed.
| What feeds the briefing | Where it comes from |
|---|---|
| Five monitored dimensions | Their existing run history — the same checks that power the dashboards and alerts, read back over time. |
| Eight researched dimensions | Compiled fresh for this edition, so the slower-moving signals are current as of when the briefing was generated. |
The two tiers, and the full list of which dimension sits in which, are on Core concepts — including the labels the product uses. This page is about what happens after the raw material is gathered.
It finds the through-line, not a stack of reports
The point of the briefing isn’t to hand you thirteen dimension write-ups to read in sequence — you already have those on the dashboards. What the synthesis adds is the read across them: it looks at every dimension, and at how each one has moved run over run, and works out the single mechanism underneath. A rank on one engine, a pricing move, a wave of new comparison pages, and a gap in your own content aren’t four separate facts — often they’re one story, and the briefing is where that story gets named.
That means it reasons in two directions at once. Across runs, it can see that a rival has owned one engine for weeks while another engine’s lead keeps rotating. Across dimensions, it can line that up with how the same rival prices and what it publishes. Put together, those become a claim you can act on rather than a pile of readings:
A pattern like that doesn’t live on any single dashboard, because no single dashboard sees more than its own dimension. The briefing is the layer that does.
A light cycle is a normal outcome, not a missing feature
Some cycles, a dimension just doesn’t have enough fresh signal to justify a full write-up — nothing funding-related happened, say, or the reviews didn’t move. When that happens, CompetLab doesn’t pad it out and it doesn’t silently drop it. The dimension is collapsed: folded into the sections it’s relevant to, with a short line explaining why it isn’t getting its own section this time.
| What you see | What it means |
|---|---|
| A dimension written up in full | It had enough fresh signal this cycle to stand on its own. |
| A dimension shown collapsed, with a one-line reason | Too little moved to warrant a full section — it’s folded into related sections, not omitted. |
This is deliberate, and it’s why every edition also carries a plain coverage note about what could and couldn’t be compiled that cycle. A collapsed dimension is CompetLab being honest that the month was quiet there — not a gap in the product, and never something to read as an empty finding.
Because the eight researched dimensions are compiled fresh each edition, they’re the ones most likely to collapse in a given cycle — a quiet month for funding or hiring simply has less to report than a busy one. The five monitored dimensions almost always have history to draw on.
Cadence, and the edition you get
A new edition is generated roughly every month. Each regeneration re-reads the monitored history and researches the eight briefing-only dimensions again, so what you open reflects everything that has shifted since the last edition.
You always read the latest finished edition — the newest complete one for your project. CompetLab keeps the current briefing rather than a shelf of past ones, so there’s no archive to page through; if a fresher edition is being generated while you’re reading, you keep the last complete one until the new one is ready. When you need to see how a single figure moved week to week, that’s what the monitored dimensions’ history is for. More on the rhythm and the reasoning is on the Strategic Briefing overview.
Work with it in code
The same edition the dashboard renders is available to your tools — including which dimensions made it into this cycle and the honest coverage caveats behind it, so a collapsed dimension reads as context rather than a hole.
FAQ
How is a briefing generated?
CompetLab compiles each edition from two streams. The five monitored dimensions already run on their own schedules, so the briefing reads their accumulated history — the same checks behind the dashboards and alerts, looked at over time. The eight researched dimensions aren't polled on a timer; CompetLab researches them fresh each time it builds an edition, so the slower-moving signals are current as of that compile. Then comes the part that makes it a briefing rather than a report bundle: it reads across all thirteen, and across how each has moved run over run, to find the single pattern underneath and turn it into a thesis, a ranked set of actions, and a competitor read. A new edition is generated roughly monthly.
What's the difference between the monitored and researched inputs?
It comes down to how each dimension reaches the briefing. The five monitored dimensions — AI Visibility, Positioning, Pricing, Content, and Tech & Trust — have been checked continuously on their own cadences, so the briefing draws on weeks of stored history and can read them as trends. The eight researched dimensions cover slower, higher-context signals like funding, hiring, and the wider market; CompetLab researches those fresh when it assembles the edition rather than keeping a live feed, so they're a periodic study rather than a running chart. Both tiers feed the same synthesis. The full list of which dimension sits in which tier, with the labels the product uses, is on the Core concepts page.
Does the briefing just stack the dimension reports together?
No — that's the difference between the briefing and the dashboards. The dashboards already give you each dimension on its own. The briefing's job is the read across them: it looks at every dimension and at how each has moved over successive runs, then works out the one mechanism that ties them together. A rank shift on one engine, a rival's pricing move, a wave of new content, and a gap in your own pages often aren't four facts — they're one story, and the briefing is where that story gets named. That cross-dimension, cross-run pattern doesn't appear on any single dashboard, because none of them sees past its own dimension. The synthesis is the whole point of the layer.
Why did a dimension show up light, or "collapsed," this cycle?
Because it didn't have enough fresh signal that cycle to earn a full section — and CompetLab tells you so rather than padding it out or quietly dropping it. When a dimension is collapsed, it's folded into the sections it's relevant to, with a short line explaining why it isn't standing on its own this time. This is normal, especially for the eight researched dimensions: a quiet month for funding or hiring simply has less to report than a busy one. Every edition also carries a plain coverage note about what could and couldn't be compiled. So a collapsed dimension is the briefing being honest about a quiet month there — not a missing feature, and never something to read as an empty finding.
How often is a new edition generated?
Roughly once a month. Each regeneration re-reads the monitored dimensions' accumulated history and researches the eight briefing-only dimensions again, so a new edition reflects everything that has shifted since the last one — independent of how often the individual monitored dimensions run. You don't request an edition; it's generated automatically for every project on that monthly rhythm. Between editions, the monitored dimensions keep updating on their own schedules and alerts keep firing, so the briefing is the periodic step back rather than your only signal. When you want to see how a single figure moved week to week, that lives in the monitored dimensions' run history, not the briefing.
Can I read the edition CompetLab is generating right now, or an older one?
You always read the latest finished edition — the newest complete one for your project. CompetLab keeps the current briefing rather than an archive of past editions, so there's no way to page back to last quarter's. If a fresher edition is being compiled while you're reading, you keep the last complete one until the new edition is ready, so you're never handed a half-written briefing. This is the same behaviour through the app, the API, and the MCP tools — the API even exposes an availability state so a script can tell a ready edition from one that's still being prepared. To track how a specific number moved over time, use the monitored dimensions' history, which is kept in full.