---
name: 01-pipeline-review
description: Run when I paste the pipeline and want the week's priorities plus the mix picture. Trigger - \"run 01\" plus the pipeline.
---

# TASK 01 — PIPELINE REVIEW

## INPUT
Deals with stage, value, and whether each is agency or brand
direct. Better: next step, owner, flight date, holdco or
independent, and last week's version for comparison.

## PROCESS
1. Rank by (value x closeability before its flight date).
   Ignore anything with no next step older than 30 days for
   the top three.
2. Compare against the previous version. Note stage or date
   slips. Flag, never scold.
3. Read the agency/brand mix against [SPLIT]. Say plainly if
   the pipeline is not pointing where the plan says.
4. Note concentration if one holdco or one brand is behind an
   outsized share.

## OUTPUT
1. PUSH THIS WEEK — three deals: account, value, the single
   next action, who owes it, by when.
2. SLIPPED — one line each: what moved and the question for
   the owner.
3. MIX — agency vs brand, actual vs intended, two lines.
4. CONCENTRATION — one line, only if there is something in it.

## RULES
- One screen. No commentary on how I run the team.
- Never estimate a deal value that was not in the paste.
- A deal without a flight date ranks below one with a date,
  whatever the size.

## EDGE CASES
- No agency/brand tags in the data: rank anyway and say the
  mix read is impossible until the tags exist.
- Everything looks healthy: say so in one line, then name the
  one deal you would still worry about and why.

## GOOD LOOKS LIKE
\"CONCENTRATION: one holdco is 61% of Q4 pipe. One procurement
policy change there is the whole quarter.\"

---
name: 02-agency-brief-response
description: Run when an agency brief or RFI lands and the response needs to go same day. Trigger - \"run 02\" plus the brief and whatever I gave you.
---

# TASK 02 — AGENCY BRIEF RESPONSE

## INPUT
The brief, or its substance: goal, audience, budget if stated,
flight, measurement. Plus anything I told you is available on
our side.

## PROCESS
1. Read the goal back in their words, one line. A wrong
   read-back kills a response faster than a wrong price.
2. Sketch the persona approach in three lines maximum, built
   only from what I gave you.
3. Ask the three questions that will shape the plan — ones
   they must answer with information, not yes or no.
4. Close with the concrete next step and a date.

## OUTPUT
- The email: read-back, approach, three questions, next step.
  120 words.
- INTERNAL: what to get from planning before any number goes
  anywhere.

## RULES
- Never quote reach, CPMs or pricing that I did not give you.
  Insert [RATE] and flag it instead.
- Never promise a custom study or a turnaround someone else
  has to deliver.
- Same-day beats complete. A sharp response with three good
  questions outruns a full plan sent Friday.

## EDGE CASES
- The brief is actually a formal RFP: answer the process — who
  owns our response, what is due when — before any selling.
- The brief is vague: the three questions become the entire
  reply, and say so to me first.

## GOOD LOOKS LIKE
\"READ-BACK: auto intenders, heavy streaming skew, Q4 flight,
measured on dealer visits. Confirm that before anything else
travels.\"

---
name: 03-differentiation-reply
description: Run when a buyer says everyone claims the same thing or asks why we deserve a line on the plan. Trigger - \"run 03\" plus what they said.
---

# TASK 03 — DIFFERENTIATION REPLY

## INPUT
Their pushback, verbatim if possible. What is already on their
plan. Who they are and how senior.

## PROCESS
1. Diagnose, one line to me: category noise, budget defence,
   or a genuine no. Only noise gets this treatment.
2. Concede what is true completely. The decks really do all
   say it. Fighting that proves their point.
3. ONE differentiator: persona targeting built from survey
   data at placement level does not degrade where identifiers
   disappear. One. Not five.
4. Hand them a test they can run without trusting us: split a
   recent campaign's delivery by where IDs were absent and
   look at what those impressions bought.
5. Close on a working session with their data, dated.

## OUTPUT
- One line to me: the diagnosis.
- The reply (OUTPUT DEFAULTS apply).
- THE RESERVE: the line for if they ask for a one-pager
  instead — held back, not sent.

## RULES
- Never rubbish their SPO partner, DSP or verification stack.
- No lift claims, no case-study numbers I did not source.
- If the honest answer is that their setup covers it, say so
  to me and do not send a differentiation email at all.

## EDGE CASES
- It is actually budget defence: differentiation reads as not
  listening. Say so and stop.
- They are hostile rather than tired: three lines maximum.
  Length reads as defensiveness.

## GOOD LOOKS LIKE
\"If your current setup held reach where the IDs weren't, you
don't need us — and I'll say so.\"

---
name: 04-cpm-pushback
description: Run when an agency or brand pushes on rate. Trigger - \"run 04\" plus their pushback and my floor if I have one.
---

# TASK 04 — CPM PUSHBACK

## INPUT
Their pushback, verbatim if possible. What they say they pay
elsewhere. My floor if I have given you one. Flight size and
term.

## PROCESS
1. Diagnose: budget ceiling, ritual negotiation, or a
   comparison mismatch. One line to me above the draft.
2. Test like-for-like. What does their comparison rate
   include — creative, data, measurement, guarantees? An
   open-exchange CPM against a managed persona campaign is
   not a comparison, it is a category error.
3. Hold with ONE reason tied to something concrete.
4. Offer one concession that is not a rate cut — flight
   timing, creative support, a learning agenda — only from
   what I have told you is available.
5. Write the walk-away line, polite, held in reserve.

## OUTPUT
- One line to me: the diagnosis.
- The reply.
- THE RESERVE: the walk-away, marked \"only if they push
  again\", not in the email body.

## RULES
- If I gave a floor, nothing below it appears anywhere, ever.
- If I gave no floor, ask for it BEFORE drafting. Hard stop.
- Never match a rate I have not seen in writing.
- Never defend the rate with a lift number I did not source.

## EDGE CASES
- The comparison is remnant or open-exchange inventory: the
  reply asks what that rate includes, and nothing else.
- Their offer is genuinely better and like-for-like: say so
  to me and tell me what we can defend on, or that we cannot.

## GOOD LOOKS LIKE
\"Diagnosis: category error — they are quoting an open-exchange
CPM with no data, no creative and no guarantee against a
managed persona flight. The reply asks what that rate buys.\"

---
name: 05-brand-meeting-prep
description: Run before a meeting with a brand, agency or holdco. Trigger - \"run 05\" plus who, what I want, and any history.
---

# TASK 05 — BRAND MEETING PREP

## INPUT
Who I am meeting and their role, the brand or agency, what I
want from the room, history if I have it. Say whether their
agency is in the room or not.

## PROCESS
1. Compress their situation from ONLY what I gave you. Mark
   anything inferred as \"guess:\".
2. Restate my objective as one sentence with a number or a
   date in it. Sharpen it if it was vague and say you did.
3. Three questions that surface signal-loss pain and how the
   plan gets made — answerable with information, not yes/no.
4. Two likely objections with one-line answers. For a brand
   with an agency, one of them is always \"talk to our
   agency\" — and the answer is yes, gladly, together.
5. Define the close: what I ask for before leaving.

## OUTPUT — one page
THEIR SITUATION (2 lines) / MY OBJECTIVE (1 sentence) /
THREE QUESTIONS / TWO OBJECTIONS + ANSWERS / THE CLOSE.

## RULES
- Never position around the incumbent agency. If the brand
  has one, the agency hears it from us first or in the same
  room. This is non-negotiable — the agencies are half the
  book.
- Nothing generic. Every question must be unusable for
  another account without editing.

## EDGE CASES
- Only a name and a title: build the page with [GAPS] and
  list the three facts to get before the room.
- The meeting is a holdco top-to-top: the close is a named
  working team and a date, never a deal.

## GOOD LOOKS LIKE
\"Q2: When your agency reports campaign reach, what share of
it still depends on an identifier being present?\"

---
name: 06-campaign-recap-email
description: Run when a flight ends and the client needs the recap. Trigger - \"run 06\" plus the campaign numbers as delivered.
---

# TASK 06 — CAMPAIGN RECAP EMAIL

## INPUT
The flight's delivered metrics, the goal they were bought
against, and anything unusual that happened mid-flight.

## PROCESS
1. Line one: the headline metric against its goal. No
   wind-up.
2. Two things that worked, at persona or placement level,
   specific enough to repeat next flight.
3. One thing to change, honest, with what we would do about
   it.
4. The next-flight ask with a date.

## OUTPUT — one email, 150 words max
THE RESULT (line one) / WORKED (2) / CHANGE (1) /
NEXT FLIGHT (the ask, dated). Subject under 7 words, never
\"Campaign recap\".

## RULES
- Only metrics I gave you. Never extrapolate, annualize, or
  convert one metric into an implied another.
- An underdelivery is stated in line one, not paragraph
  three. Burying it costs the renewal.
- A make-good is offered only if I said so.

## EDGE CASES
- The flight underdelivered: same structure. Line one is the
  gap, the cause is named, and the change section is the
  whole email's weight.
- Results are great but the sample is thin: say so rather
  than headline it. Agencies keep receipts.

## GOOD LOOKS LIKE
\"94% on-target reach against a 78% plan benchmark — and the
two placements that dragged it are already cut from the next
flight.\"

---
name: 07-week-plan
description: Run Sunday night or Monday morning. Trigger - \"run 07\" plus calendar, live deals, team matters, anything on fire. Reads on a phone.
---

# TASK 07 — WEEK PLAN

## INPUT
Calendar, live deals, team matters, anything on fire.

## PROCESS
1. Three OUTCOMES, one from each: agencies, brands, team. If
   one side has nothing, name it and say why that is the
   risk.
2. What to decline or delegate, with a one-line script so it
   costs me nothing to send.
3. The thing I am avoiding. Derive it from what I pasted —
   in a new role it is usually a relationship not yet opened,
   not a deal.

## OUTPUT — half a page, phone-readable
THREE OUTCOMES (agencies / brands / team) /
DECLINE OR DELEGATE (item + script) /
THE AVOIDED THING (one line).

## RULES
- Outcomes state what is TRUE on Friday, not what I will do.
- Maximum three. Say what you parked and why.
- In the first quarter of a new role, an external meeting
  outranks an internal one in a collision, and say so when
  they collide.

## EDGE CASES
- Week is wall-to-wall internal: the first outcome becomes
  carving external time, with the two meetings to move named.
- Travel week: the team outcome becomes one decision made,
  not presence.

## GOOD LOOKS LIKE
\"AVOIDING: the holdco you haven't met since the announcement.
Every week that passes prices your intro meeting down.\"

---
name: 08-ceo-one-pager
description: Run at month end for the CEO. Trigger - \"run 08\" plus the month's numbers and anything he should hear from me first.
---

# TASK 08 — CEO ONE-PAGER

## INPUT
Americas revenue vs plan, pipeline, wins and losses with
values, team changes, and anything Nicolas should hear from
me before he hears it elsewhere.

## PROCESS
1. Line one is the Americas number vs plan. No wind-up.
2. Three wins with values and whether agency or brand. Two
   losses with the honest reason — price, process, product
   or relationship, never merged into \"competitive
   pressure\".
3. Pipeline health in two lines: coverage and the biggest
   risk.
4. Team in one line, only if something moved.
5. The ONE decision or support I need, with a date.

## OUTPUT — one page, in this order
THE NUMBER / WINS (3) / LOSSES (2, honest) /
PIPELINE (2 lines) / TEAM (1 line, optional) /
THE ASK (1, dated).

## RULES
- No adjective a number could replace.
- Losses name the reason, never the seller.
- A bad month gets the same structure and length. The
  temptation to write more is the tell.
- Never speculate about other regions. Americas only.

## EDGE CASES
- Numbers moved for a market reason outside our control: say
  so. Taking credit for luck costs the next conversation.
- No ask: say \"No decision needed this month\". Never invent
  one to fill the slot.

## GOOD LOOKS LIKE
\"LOST: $410k CPG brand — their procurement required an
incumbency waiver we couldn't clear inside their window.
Process, not product.\"