Where we start
Partnership session 2 tagged every ask and wrote a real C for each split one. That works when an analyst has time to think about each ask. In a busy month nobody does, so the thinking has to be built into the front door: a short form that every ask passes through, and a menu that says, before anyone asks, what data does and how long it takes. The running company is Kettle & Co, constructed; the turnaround times below come from the triage desk's modelled hours and will differ for your team.
★ What you walk out with tonight. An intake template you can paste into any ticket tool, a one-page service menu with three turnaround bands and a "not on the menu" list, and one worked example of a vague market research ask turned into a three-day deliverable that leaves the research with its owner.
A front door makes every ask smaller
4 min liveMost asks are bigger when they arrive than when they are understood. "Can you do market research on Japan?" sounds like forty hours. Asked what decision it serves, who decides and by when, it becomes "the BD lead needs to know what our own orders already say about Japan before she briefs an agency next week", which is three. The front door is where that shrinking happens, every time, without depending on the analyst's mood that day.
LiveWhat the front door is, and what it is not3 min▶
It is one short form, the same for every requester, that the data team fills in with the requester in a ten-minute conversation, or that the requester fills in alone for small asks. It is where the knowledge tag from Partnership session 2 gets decided, and where a data C gets its deliverable and date.
It is not a bureaucratic gate. If the form takes longer than the work, the form is wrong. Small standard asks from the menu (an email send read, a before and after on a PR hit) should go straight through on a one-line request. The door is for asks that are vague, big, or not obviously data's.
Running the conversation itself, the discovery questions and how to listen for the real need, is taught in Asking the Right Questions as a DA ↗. This page is the form that conversation fills.
Self-studyWhy the requester should own the form, not the analyst2 min▶
A form the analyst fills in after the meeting is the analyst's interpretation. A form the requester signs is an agreement. The difference shows up three weeks later, when the requester remembers asking for something bigger: a form in their words, with their decision and their date, settles it without anybody having to be right.
The inbox that became a queue. A data team that took asks over chat, email and hallway conversations could never say how many asks it had, let alone which were late. One shared form, with five required fields, turned the same volume into a queue anyone could read, and the first month's surprise was that a third of the asks were the same four standard reads. Those went onto the menu. Constructed example; the finding is a common one.
The intake template
6 min liveFive fields the requester owns, then three the data team adds. The first five are the four questions from Angles session 1 with the business question in front. The last three are the ownership map from Partnership session 2, applied to this one ask.
LiveThe template, ready to paste4 min▶
Copy this into your ticket tool, a shared document, or a form. Keep it to these eight fields; every extra field is a reason not to use it.
Field 5 is the one people skip and the one that matters most. If both answers lead to the same action, the analysis is not needed for this decision, and the honest move is to say so and offer what would change the action instead.
Self-studyThree ways the template gets gamed, and the fix for each3 min▶
"Decision: to understand the market." Understanding is not a decision. Ask what they will do once they understand it.
"Deadline: ASAP." ASAP has no date. Ask when the meeting is where the decision gets made.
"Decision owner: data." Then nobody in the business will act on it. The owner is the person who will be judged on the outcome; data informs the bet and the business places it.
A one-page menu of what data offers
5 min liveA menu does two jobs. It tells marketing, brand and BD what they can ask for without negotiating, and how long each thing takes. And it says, on the same page, what is not on it and who owns that instead, which is the kindest form a boundary can take: written down before anybody asked.
LiveThe menu, ready to adapt3 min▶
Turnaround comes from the triage desk's modelled hours: at six request hours a working day, 2 hours or less is one working day, 3 to 6 hours is three working days, 8 to 10 hours is one to two weeks, and anything bigger is a project for the monthly plan. Your hours and days will differ; the bands are the idea.
| On the menu | For | Turnaround |
|---|---|---|
| Email send read: clicks and orders per send, opens set aside | Marketing lead | 1 working day |
| PR before and after: branded search and direct visits against last year | Brand manager | 3 working days |
| Query-to-revenue join: search queries to landing pages to orders | Marketing lead | 3 working days |
| Campaign readout against the pre-agreed metric | Marketing lead | 3 working days |
| Holdout design, and the read after the test runs | Marketing lead | 3 working days, plus the test |
| Creator, affiliate or event sales table: orders and new customers by code, link or scan | Marketing, BD | 3 working days |
| Evidence page for market research: what our own data already says | BD lead | 3 working days |
| Bottom-up half of a market sizing (method: Market Sizing, Analyst 2 ↗), or the account evidence for an ICP | BD lead | 3 working days |
| Tracker wave read: weighting and significance | Brand manager | 3 working days |
| CAC, margin and payback by acquisition channel | Marketing lead, finance | 1 to 2 weeks |
| A weekly dashboard: one decision per tile, definitions under every number | Marketing lead | 1 to 2 weeks |
| A subscription or sales dip, explained by channel and tenure | Marketing lead | 1 to 2 weeks |
| Not on the menu | Who owns it | What data can add, if asked |
|---|---|---|
| Market research fieldwork, surveys, panels | BD or brand, with an agency | The evidence page above |
| Market penetration or go-to-market strategy | BD and marketing | Customer evidence by region or segment |
| Competitor reviews | Brand | Our own price-change history |
| Agency or influencer negotiations | Brand, procurement | Last quarter's partner sales table |
| Creative and positioning briefs | Brand | Nothing, usually; say so kindly |
Self-studyKeeping the menu honest2 min▶
A menu with turnaround times is a promise. Track how often each item lands on time for a quarter, and change the time on the menu rather than missing it quietly. Add an item when the same ask arrives three times in a month; remove one nobody has ordered in a quarter. And when capacity drops, as it did on the triage desk's break button, change the turnaround bands that month and tell everyone the same day.
The "not on the menu" list that ended an argument. A BD lead and a data lead had the same argument every quarter about who should do market research. Once the menu said, in a printed list both had agreed, that fieldwork sits with BD and an agency and that data supplies an evidence page in three days, the argument became a two-line request. Constructed example; the mechanism is writing the boundary down before the ask.
The Japan ask, through the front door
★ 22 min · everyone works it throughNo code. Kettle & Co's BD lead sends ask #11 from the triage desk on working day 8: "Can data do the market research on matcha drinkers in Japan?" Run it through the intake template, field by field, then match it to the menu and write the confirmation note. Then run one of your own asks the same way. Every Kettle & Co value is constructed; yours will differ.
Step 1 · Business question and decision. Ask the BD lead what she will decide. At Kettle & Co: whether to enter Japan next year. Write it as two possible actions.
Step 2 · Owner and deadline. The owner carries the outcome. The deadline is the decision date, here the agency brief, not "ASAP".
Step 3 · What changes. Make her say what she would do on each answer. If both answers lead to the same action, stop and say so.
Step 4 · Tag it. Mostly external: Japanese buyers, distributors and culture live outside our data. One internal thread: what our own orders and search data already say.
Step 5 · Match the menu. "Evidence page for market research", three working days. Lands day 8, due day 15: it fits. Write what is not data's in the same field, with its owner.
Step 6 · Send the confirmation note. Two sentences, deliverable and date first, then the boundary with its reason. Partnership session 4 is all about this note.
Print the template and the menu together. Expand all, then print this page, or paste both into the place your requesters already work. A front door nobody can find is a wall.
Before Partnership session 4
30 min- Put the template somewhere real: your ticket tool, a form, or a pinned document, and send the link to one requester.
- Draft your own menu from your last month of asks: the standard reads, their real turnaround, and a "not on the menu" list with owners.
- Run two of your own asks through the template, and note which field the requester found hardest.
- Bring the hardest complaint you hear about the menu or the boundary. Partnership session 4 writes the script for it.
Three questions
Clear all three and Partnership session 3 gets stamped in your passport.
1 · A requester fills in "Decision: to understand the market better". What should happen next?
Understanding is not a decision. The decision names the measure, the precision and the date; without it the analysis has no edge.
2 · Why does the service menu include a "not on the menu" list with owners?
A boundary written in advance is the kindest kind. The "what data can add" column keeps every split ask's evidence half alive.
3 · The Japan ask arrives as "can you do market research?". After intake, what does data deliver?
Three modelled hours instead of forty, and the decision stays with the person who will be judged on it.
What this session teaches, and where it came from
Every fact on this page, and its verification tier, is recorded in the course's source map.
Partnership session 3 cheat sheet · pin this
You have a front door and a menu. Partnership session 4 is the conversation when somebody pushes on them: scripts for the six loudest complaints, each opening with a deliverable and a date before the boundary, the escalation path, and when the right answer is simply yes.