How Pre-Seed Founders Book 10 Design Partner Intros in 14 Days
Design partners are the cheapest pre-seed funding round you'll ever raise. Ten of the right ones give you product validation, paid revenue, and three logos on your deck. Ten of the wrong ones give you a roadmap full of features you can't sell to anyone else. Most founders take 90 days to land 10 partners and end up with the wrong mix. This is the 14-day version, run cold, run from a clean LinkedIn account, with a binary commercial close on day 14.
The routine assumes a sharp problem hypothesis, a 30-second demo of some kind (even Figma frames are enough), and a LinkedIn account you log into daily. It does not assume a network of warm intros or a list of VCs you can lean on. The total cost is roughly 480 LinkFetch credits across the 14 days, which at flat per-call pricing comes in well under a single Sales Navigator seat.
Why cold beats warm for design partner outreach
The default founder advice is to ask your VC for design-partner intros. It feels safer. It also degrades the signal you came here for. A partner who said yes because your investor asked nicely is a partner who'll churn the day your investor stops asking. Strella's co-founders recruited all 12 of their design partners through cold LinkedIn outreach with no warm intros, then converted 12 of 12 to paid, because the harder it was to recruit them, the stronger the buying signal (Bessemer Venture Partners, 2026).
Cold also widens the universe. Your 1st-degree network optimizes for who you already know, not who has the problem. The 14-day routine inverts that: search the full LinkedIn graph for the operator role inside your ICP, then use relationship proximity as a tie-breaker on otherwise tied candidates.
The third reason cold wins: it forces you to write a pitch that holds up without social proof. If your DM lands without an investor name attached, your messaging is shippable to anyone. If it doesn't, you found a positioning bug before you wasted three quarters selling.
The ICP filter that actually converts
The single biggest mistake at this stage is picking design partners by company logo recognizability. Stripe and Shopify alumni look great in your deck and will ghost you on the third week, because big-company operators have no decision power on tooling adopted post-hire. You want the opposite: the person who can sign a $20k pilot in their first month without asking permission.
Three filters, applied in order, narrow a 5,000-person LinkedIn pool to a 200-person target list:
- Role with budget. Heads of, VPs, founders, or director-level operators in companies of 20 to 200 people. Not C-suite at F500. Not ICs anywhere.
- In-seat 30 to 240 days. New-in-seat operators are buying software to prove themselves in their first quarter. After 240 days they've already picked their stack and won't replace it for six months. This is the highest-converting single signal you have, and it maps directly onto the new-hire tracking recipe.
- Active on LinkedIn. At least one post or comment in the trailing 30 days. A profile that hasn't been touched in a year means your DM lands in an inbox no one reads.
Of the 200 who survive, the a16z framework recommends focusing on operators whose pain you can prove from their public footprint, not just whose title matches your ICP (a16z, 2026). The Claude scoring step in day 4 does that proof step automatically.
Day 1 to 3: build the candidate universe
The goal of the first 72 hours is to get a clean 200-person list into a spreadsheet with one row per candidate and seven columns of context. Not 2,000. Not 50. Two hundred.
Open Claude Desktop, attach the LinkFetch MCP, and run the source prompt:
Search LinkFetch for profiles where headline contains "Head of Revenue Operations" OR "VP RevOps" OR "Director of Sales Operations", current company size between 20 and 200, current tenure between 30 and 240 days, and country in US, UK, DE, NL. Return 250 profiles with name, headline, company, company size, tenure, profile URL, and the last public post URL.
That call costs roughly 250 profile credits plus 250 company credits, so 500 credits total in one shot. From the 250 returned, trim by hand to 200, dropping anyone whose company is on a LinkFetch known-Notion-customer or known-HubSpot-customer list if your competitor is one of those. The hand-trim takes 30 minutes and is the single highest-leverage step in the routine.
By end of day 3 you should have a clean 200-row sheet, the
candidate's most recent post URL, and a column called pain_hook
that's still empty. Day 4 fills it.
Day 4 to 7: score and personalize with Claude
The pain_hook column is the entire reason your reply rate goes from 6% to 22%. It's a one-sentence summary of why this specific operator should care about your product, written in their language, sourced from their last six months of public activity.
Run this Claude prompt over the 200 rows in batches of 20:
For each profile below, fetch their last 10 LinkedIn posts via LinkFetch, plus the 3 most recent posts they reacted to. Identify the single pain point they mentioned most often. Write a one-sentence pain_hook in their tone, max 18 words, no jargon, no "as a [title]" openers. Output as CSV with profile_url, pain_hook.
That's another 200 profile-activity calls, roughly 200 credits. The output is uneven. Maybe 40% of rows come back with a sharp, specific hook ("you said last week that pipeline forecasts break when reps mis-attribute closed-lost reasons"). 40% come back generic and useless. 20% come back with no signal because the operator posted three times about their dog. Drop the bottom 60% and keep the top 80 rows.
Day 7 ends with 80 rows in your sheet, each with a tight pain_hook. This is the universe you'll actually outreach.
Day 8 to 12: the cold DM that converts at 22%
Most founder DMs fail on the first line. The opener is some variant of "Saw you're VP RevOps at Acme, would love to chat about our AI tool." It reads as automated even when it isn't, and the recipient is gone before paragraph two.
The 22%-reply DM is built in three parts:
Line 1: Reference their pain_hook in their own words.
Line 2: One sentence on why your product addresses it,
no product noun, no buzzwords.
Line 3: Binary ask. "Want to be one of 10 design partners?
It's free, ends in 90 days, you get a 50% lifetime
discount if it works."
The binary ask is critical. Asking "would love to chat" gets 4% replies. Asking "want to be one of 10 design partners with these specific terms" gets 22%, because you're respecting their time and proving you've thought about the program structure (Unusual Ventures, 2026).
Send 10 DMs a day for five days, totalling 50 sends across the 80 rows. The rest stay in reserve for round two. Track replies in a second sheet column. By end of day 12 you should have 10 to 12 positive replies, 3 to 5 "interesting, talk later" replies, and the rest no-answer.
Mid-week the routine gets dull. Most founders quit on day 9 because they sent 30 DMs and got 4 replies and feel like the whole thing is broken. It isn't. Replies cluster late in the second week as recipients work through backlog. Keep sendig the 10-a-day until day 12.
Day 13 to 14: book the intros and ask the binary
By day 13 the calendar should have 10 intro calls slotted into the following week. Each one runs 30 minutes, structured as:
- 5 minutes: their context, their pain, no demo
- 15 minutes: demo the wedge, ask them to interrupt
- 10 minutes: the binary close
The binary close is the part founders skip. The script is roughly: "Here's the program. 90 days. Weekly call. You pay $5k upfront, we refund 100% if it doesn't ship. At day 90 you either commit to the paid contract at the discount or we part as friends. Are you in?"
This is the part Strella nailed: a hard deadline plus a binary commercial ask, applied to every partner. When all 12 said yes, that was their PMF evidence (Bessemer, 2026). Six to eight of your 10 should convert. If fewer than four convert, your product hypothesis is wrong and you should rewrite it before sending another DM.
What to track in week 1 of the partnership
Once you have 10 paid design partners, the trap is treating them like customers. They're not. They're product co-conspirators with a paid relationship, and the reporting cadence should reflect that.
The week-1 setup runs on three Claude routines, all wired through LinkFetch:
| Cadence | What it does | Credit cost |
|---|---|---|
| Daily 09:00 | Pull each partner's LinkedIn activity, flag any post mentioning the problem area | ~30 credits/week |
| Weekly Monday | Summarize the 5 most useful product complaints across the 10 partners | ~0 LinkFetch credits (Claude only) |
| Bi-weekly Friday | Track whether any partner has changed roles or company, since that's an instant retention risk | ~20 credits/week |
The first routine is the cheapest signal source and the highest ROI. When a design partner posts about a workflow problem, that's the exact feature you should ship next week. The third routine is the unglamorous one that saves you from a partner ghosting because they quit their job in week 4 and forgot to tell you.
For the full daily-summary cadence including the Claude Desktop MCP config, see the solo-founder beta-user routine, which uses the same loop with a different target list. For the month-2 outreach to investors based on the same dataset, the investor outreach list playbook applies the same filter-rank-personalize pattern.
FAQ
How many design partners is the right number for a pre-seed company?
Between five and twelve. Below five you can't see patterns in the feedback. Above twelve you can't give each one enough founder time to keep the partnership warm, and the program loses its high-touch character. Ten is the sweet spot, and ten is also what the binary ask scales to without burning your week. If you can only land six, that's still a working program.
Should I pay design partners or charge them?
Charge them, ideally $5k to $20k upfront with a 100% refund clause if you fail to ship. Free design partners churn at 60%, paid ones at 10%. The money matters less than the commitment device. If the partner can't sign a $5k check, they can't sign your $50k contract in nine months either, and you've selected for the wrong buyer.
What if my product doesn't have a demo yet?
You can run the routine with Figma frames and a Loom walkthrough, but cap the call at 20 minutes and replace the demo with the problem-validation script. Use the call to confirm the pain is real and the buyer has budget. Skip the binary close to day 60 instead of day 14, after the first real build. The candidate sourcing work is the same either way.
How is this different from beta-user recruiting?
Beta users test usability and don't pay. Design partners commit to a 90-day paid pilot with a weekly call and a roadmap influence seat. Beta is a $0 signal at 38% reply rate. Design-partner is a $5k signal at 22% reply rate. Do both: beta in month 1 to harden the build, design partners in month 2 to validate the commercial model. The solo-founder beta-user routine covers the first half.
What does LinkFetch charge for this routine end to end?
Roughly 480 credits across the 14 days: 250 for the candidate search, 200 for the personalization activity scrape, 30 for the week-1 partner monitoring setup. At flat per-call pricing that's a single sub-$50 invoice for a process that normally takes three months of founder time. If you re-run the source step at day 30 because the first cohort was too narrow, budget another 250.
Last updated 2026-06-22 by the LinkFetch team.