Prompt 08 — Build the list of 100

Read AGENTS.md first, then stack/CONTEXT.md, stack/02-customer/customer.md,
stack/03-market/market.md and stack/05-gtm/gtm.md.

## Step 0 — Check the ground

Open stack/02-customer/customer.md. If it is still [TBD], stop.

A hundred names built against a customer nobody has defined is a hundred wasted
approaches, and the founder will conclude the channel does not work rather than
that the targeting was invented. 03-customer-and-problem.md fills it.

A thin or empty stack/05-gtm/gtm.md is not a problem here. It is the file this
prompt writes.

## Step 1 — Pick the channel

Look at every channel in 05-gtm and tell me:

- Which have been genuinely tested, with what result
- Which are untested
- Which ONE you would work first, and why

Weight your recommendation toward channels the founder can physically walk into
this week — their own neighbourhood, their last employer's industry, the
community they already belong to. Reachability beats attractiveness. Founders
systematically choose a channel they find impressive and cannot access.

Ask me to confirm the channel before continuing.

## Step 2 — Define what "yes" means

For this channel, specifically:

- What is a win? Stocked 5 units / signed a pilot / placed a trial order / took
  a meeting / made an introduction
- What is the realistic conversion rate? (Roughly 5% is a common working
  assumption for cold outreach into a warm-ish local channel. If 100 produces
  zero, the problem is the offer, not the list.)
- How many wins does the founder need for this channel to be worth continuing?

## Step 3 — Build the list

Produce as many named, specific prospects as you can — target 100. For each:

  Name | Organisation | Location | Contact person and role | Phone | Email |
  Website or socials | Source | Segment | Why they fit

Use several sources; one source runs out at forty:
- Maps and local search, worked systematically by area
- Industry directories and association member lists
- Competitors' visible customer lists
- Alumni networks, university offices, chambers of commerce, professional bodies
- Referrals from existing customers
- Trade show and exhibition attendee lists
- Public tender and procurement portals
- Job-title search on professional networks
- Simply walking a neighbourhood and writing down every one

CRITICAL: only list organisations you can actually verify exist. Do NOT generate
plausible-sounding business names to fill the table. A list of 40 real names is
worth infinitely more than 100 with 60 invented, and inventing them destroys
trust in everything else in this stack.

If you can only verify 40, deliver 40 and tell me exactly where to find the
other 60 — which directory, which search, which office to ask.

## Step 4 — Introductions to ask for

Who could open doors here? Alumni offices, incubators, chambers of commerce,
existing customers, mentors, local associations.

Draft the actual ask. It must be specific and small:

  "Could you give me two or three names of offices here I could approach?"

not

  "Do you know anyone who might be interested?"

The second gets nothing. Also note: when asking a favour, one product nicely
presented plus a business card goes a long way. Not the whole range.

## Step 5 — The cadence

Write out the sequence, with the actual message for each step:

  Day 0   Email or message — who you are, what it is, the order sheet or link.
          Make saying yes trivially easy.
  Day 7   Follow-up. A different angle, not a repeat.
  Day 14  Second follow-up.
  Day 21  Phone call.
  Day 28  Visit in person.

Four follow-ups feels excessive and is not. Most people who eventually say yes
did not reply to the first message. Founders stop at two and conclude the
channel does not work.

Draft each message. Short, specific, one clear ask, easy to say yes to. Use the
customer's own language from 02-customer, and in a local market draft it in the
local language as well as English — local-language outreach routinely
outperforms and founders default to English because that is what they write
proposals in.

Set a daily target: how many outbound touches per working day. Ten a day is 200
a month.

## Step 6 — The in-person pitch

Thirty seconds, said out loud, then a specific ask. Draft it for this channel.
It should:

- Say who they are and that they are local
- Say what it is in one line, in the customer's words
- Remove all risk: small quantity, free or on consignment, short window, "I'll
  come back either way"
- Make reordering trivial — a QR that opens a pre-drafted WhatsApp message
- Ask for information as well as a sale: what do you think would sell, what
  would you price it at, what do you stock in this category already

## Step 7 — What to take

List the collateral needed before the first visit, and flag anything missing
from 05-gtm:

- Customer flyer (A5, glossy, visual, benefit-led)
- Trade or order sheet (price, THEIR MARGIN, minimum order, delivery time,
  payment terms, WhatsApp QR) — see worksheets/trade-order-sheet.md
- Business card
- Samples
- Barcode, if the channel is organised retail

A shopkeeper cannot stock you if they do not know their margin. If the trade
sheet does not exist, that is the blocker — building it comes before the visits.

## Step 8 — Tracking

Define the sheet: columns, statuses, and the weekly numbers to report in the
recap (list size, contacted, followed up, called, visited, meetings, wins,
objections logged).

## Output

Save the list where I tell you. Update stack/05-gtm/gtm.md with the channel, the
target, the cadence and the collateral gaps. Update front matter and the INDEX
row.

Finish by telling me the three things most likely to make this channel fail, and
what would tell us early.