Strategy Desk

Campaign Strategy & Briefs

For: Brand managers, agency strategists, and campaign leads on both sides of the client-agency table

Creative brief builder: from messy kickoff notes to one page

Kickoff happened, the notes are chaos, and creative starts Monday. Get the single-page brief that prevents three rounds of rework.

Copy & paste into your own LLM · open in archive →

You are a senior brand strategist writing a creative brief. I will paste my kickoff notes and background material. Produce:

A) A ONE-PAGE BRIEF with labeled fields: business objective (with the number it is supposed to move), audience (described as observed behavior, not demographics alone), the single-minded proposition (ONE sentence — if it needs a comma to hold two ideas, it is two propositions), reasons to believe (drawn only from the material I provided), mandatories and constraints, deliverables and channels, and the KPI with its measurement window.
B) A GAP LIST — every brief field that could not be filled from my notes, each phrased as the exact question to ask the stakeholder, not a vague 'clarify audience'.
C) A 3-line CREATIVE SPRINGBOARD — tensions or insights actually present in my notes worth exploring, each labeled a hypothesis.

My material: [PASTE: kickoff notes, background docs, objectives, audience info, constraints, deadlines]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why it works: Most creative rework traces to a brief that buried three propositions in one sentence or never defined success. Forcing a single-minded proposition plus an explicit gap list turns brief-writing from a formatting exercise into a decision checkpoint — the questions surface before creative burns a week on the wrong read.

Useful? 👍 Yes👎 No

Copy Desk

Content & Copy

For: Content marketers, copywriters, and the generalists who ship words every day

Brand-voice-locked rewrite with a change log

The draft is fine but it does not sound like the brand — and 'make it more on-voice' is not feedback anyone can act on. Make the edit auditable.

Copy & paste into your own LLM · open in archive →

You are a brand editor enforcing a voice guide. I will paste our voice guidelines and a draft. Produce:

A) The REWRITE — preserving every fact, claim, offer, and number exactly as given. Do not add, soften, strengthen, or invent claims; your job is voice, not substance.
B) A CHANGE LOG table: original phrase, revised phrase, and the specific voice rule that triggered the change. If you cannot cite a rule for an edit, do not make the edit.
C) VOICE GAPS: every place the guidelines gave you no guidance, each phrased as a question the team should settle and add to the guide.

My inputs: [PASTE: voice guidelines (traits, do/don't lists, example passages) + the draft to rewrite + where it will run]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why it works: 'Doesn't sound like us' becomes fixable the moment every edit cites the rule behind it. The change log doubles as voice-guide QA — the recurring gap questions are exactly the rules your guidelines are missing, so the document improves every time the prompt runs.

Useful? 👍 Yes👎 No

Performance Desk

Paid Media & Attribution

For: Performance marketers, media buyers, and the agency leads who have to defend the numbers

Weekly performance read: verdict first, story second

Monday morning, accounts are live, and someone will ask 'how's it going?' by 10am. Turn the exports into a verdict-first read that survives follow-ups.

Copy & paste into your own LLM · open in archive →

You are a performance marketing analyst writing the weekly account read. I will paste this period's metrics and context. Produce:

A) A VERDICT — on track, drifting, or off track versus the targets I gave — with the supporting arithmetic shown from my numbers only.
B) A MOVERS TABLE: the biggest changes versus the prior period, each with the most defensible explanation available IN MY DATA — and the honest entry "cause not visible in this data" where that is the truth, plus the cut that would reveal it.
C) ACTIONS: three changes ranked by expected effect, each labeled reversible or hard-to-reverse, plus the single metric to watch as this week's leading indicator.

My data: [PASTE: spend, impressions, clicks, conversions, CPA/ROAS by campaign or channel — this period and prior — plus targets and any known external events]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why it works: Weekly reads either drown in screenshots or assert causes the data cannot support. Permitting 'cause not visible in this data' is the honesty valve — it routes energy into pulling the missing cut instead of into a confident story that unravels in the client meeting.

Useful? 👍 Yes👎 No

Lifecycle Desk

Lifecycle & CRM

For: Email and CRM marketers, lifecycle managers, and retention owners

Lifecycle leak-finder: where the funnel actually loses people

You own 'retention' in the broadest sense. Map the lifecycle against your real numbers and find the one leak worth a quarter of work.

Copy & paste into your own LLM · open in archive →

You are a lifecycle strategist finding the highest-value leak. I will paste my stage numbers and current coverage. Produce:

A) A STAGE MAP table: lifecycle stage (as I define it — acquisition, activation, repeat, retention, win-back), my metric and number for it, the flow or program currently covering it (or "UNCOVERED"), and the leak size in ABSOLUTE volume computed from my numbers.
B) THE ONE LEAK to prioritize, argued in arithmetic: largest absolute recoverable volume, not largest percentage — show the comparison that eliminates the runners-up.
C) An INTERVENTION SHORTLIST for that leak: three options with effort level, the mechanism by which each would work, and the measurement — including a holdout — that would prove it did.

My numbers: [PASTE: stage definitions and counts (signups, activated, repeat, churned), current flows per stage, timeframe]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why it works: Lifecycle teams sprinkle effort across every stage because everything looks improvable in percentages. Ranking leaks by absolute recoverable volume — and exposing the stages with no flow at all — points a quarter of work at the place the arithmetic says it pays.

Useful? 👍 Yes👎 No

Reporting Desk

Analytics & Reporting

For: Marketing managers and analysts who report upward, or agency leads who report to clients

The Friday report: headline-first, noise-filtered

Same ritual every week: pull the exports, rebuild the deck, guess at the story. Turn the metrics dump into a headline-first report in one pass.

Copy & paste into your own LLM · open in archive →

You are a marketing analyst writing the weekly report. I will paste this week's metrics, history, and targets. Produce:

A) THE HEADLINE — one sentence carrying the single most decision-relevant fact of the week, with its number.
B) THE REPORT in three short sections: performance versus target (arithmetic from my numbers), notable changes with the best-supported explanation available in my data, and actions in motion. Movements inside the normal range I provided get labeled "within normal variation" and are NOT narrated; if I gave you no historical range, write "variation range needed" and say how to compute it.
C) THE PARKING LOT: questions this week's data raises but cannot answer, each with the specific cut or report that would answer it.

My data: [PASTE: this week's metrics, prior weeks for range, targets, anything known about external events]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why it works: Weekly reporting burns hours narrating noise — every two-percent wiggle gets a story and readers learn to skim. A report that is allowed to say 'within normal variation' earns trust for the weeks the headline actually matters, and the parking lot turns reporting into an analysis backlog instead of a dead end.

Useful? 👍 Yes👎 No

Which section is most useful to you this week?

One tap — it shapes which prompt leads next week.

What would you like to see more of?

Prompt of the Week (Pro)

This week's bonus: a 'creative fatigue early-warning' prompt that turns your ad-level frequency and click-decay data into a swap schedule — every running asset ranked by remaining runway, with the replacement brief pre-drafted. Pro members get the full prompt in the archive.

Your searchable archive

Every prompt we’ve shipped, organized by section and task.