The problem this solves
Launch weeks compress a quarter's worth of marketing consequence into days, and the information systems around them stay on their monthly rhythm. The channel fills with screenshots and "anyone know the lead count?"; whoever can pull numbers becomes the bottleneck; and the one anomaly that mattered - the form quietly erroring, the campaign burning budget on the wrong audience - surfaces on Thursday instead of Tuesday morning.
War rooms fail in two directions: starved of numbers, they run on vibes; flooded with dashboards, they stop reading them by day two. The working format is a pulse - short, scheduled, identical in shape - so the room absorbs the state in thirty seconds and anomalies stand out precisely because everything else looks like yesterday.
This mission is that pulse. Twice a day during the flight it pulls new leads with quality scores, conversion-event acceptance, and spend, compresses them into a fixed short format, flags anything that deviates from the flight's own baseline, and posts to the war-room channel. The bottleneck human returns to their actual job.
How the mission runs
- Snapshot the funnel. Each pulse pulls the deltas since the last one: new form submissions with scores, conversion events accepted by LinkedIn, spend by campaign - the whole captured-to-streamed funnel in one read.
- Compare against the flight baseline. Numbers are read against this launch's own rhythm - morning versus evening, day over day - so "anomalous" means deviating from this flight rather than from a generic norm.
- Flag what deviates. A lead-quality collapse, a campaign's spend spike, an event batch suddenly rejected - deviations get one plain sentence each with the number behind it. Silence about a metric means it is on script.
- Compose the standing-height read. The pulse is formatted to a fixed shape: leads, quality, conversions, spend, flags. Same order every time, short enough to read standing up, so day three's pulse takes ten seconds because day two's taught the shape.
- Post on the war-room cadence. Twice daily to the launch channel by default - morning and end-of-day - with the cadence tunable to the flight. Replies in the thread become investigations with the pulse as shared context.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A war room with a heartbeat: twice-daily pulses of leads, quality, conversion acceptance, and spend in an identical short format, anomalies flagged against the flight's own baseline - the launch team steers from shared numbers while the person who used to compile them does their real job.
Make it yours
- Tighten to every three hours for one-day moments - a webinar, a product-of-the-day run - where the flight is measured in hours.
- Add a closing-day retro pulse that rolls the whole flight up: totals, quality mix, cost per qualified outcome, and the three lessons.
- Mirror a leadership edition to the executive channel at half the cadence with the same numbers and none of the operational flags.
Frequently asked questions
Why a pulse instead of a live dashboard?
Dashboards require someone to look; pulses arrive. During a launch the difference is everything - the channel gets the state pushed on a rhythm, the shape stays identical so deviation is instantly visible, and the thread under each pulse holds the discussion a dashboard never captures.
What counts as an anomaly?
Deviation from the flight's own baseline: quality mix sliding against yesterday, a campaign spending ahead of its share, conversion batches rejecting where they accepted before. Each flag carries its number, and a pulse with no flags says so in one word.
Does the war room see raw lead details?
The pulse reports counts, scores, and movements; the leads themselves flow through the CRM and speed-to-lead playbooks. Channel members see the state of the flight; person-level records stay where records belong.