Thesis and the three bets
On this page
Maktura gets everyone in a business visible on LinkedIn, in their own voice, inside the firm's rules, with Holly doing the heavy lifting. The old demo moment (a named list of everyone who engaged, pushed to a CRM) is, on our reading of LinkedIn's developer terms (research, not legal advice; D1 open), not allowed. The less exciting parts (per-person voice, policy rules, approval chains, an activity log, a colleague that asks before it acts) are the hardest to copy and the safest to say aloud. UK professional firms buy protection from embarrassment as much as visibility. A small team cannot outspend Hootsuite or undercut Paiger. It can be the vendor whose every claim, counter, deadline and data path survives scrutiny. It cannot sell at volume, so this plan rations founder time.
- Holly-led, governed own-voice visibility, sold through the trust ladder (Observe, Suggest, Auto-draft, Auto-publish, per person). Engagement is seen and replied to by the post's owner, never exported, until D1 clears.
- Founder-network, human-assisted selling into recruitment and law in parallel, paid from day one. No free trial; a named person on every paying account.
- Honest founding bands, then expansion inside accounts. Spark £69, Ignition £89, Momentum £120, Velocity £150, then Standard £250 stepping down to £100 at 400 or more licences. The rate holds while subscribed, so cohort growth comes only from added seats. The real pricing question is the 1 January step to £250 (D28).
Lessons we are applying. Sell one job and ship the platform: "my people stopped posting and I can't make them start without risking what they say". Onboarding shows voice, Holly drafting from something real, approvals and the calendar; Watch, campaigns and deeper reports come after activation. Human-assisted conversion beats a free funnel, so no freemium core and payment at signup. Report paid and posting seats, never sign-ups. Seats are the only source of net revenue retention above 100% on a cohort whose rate holds. Lead with the person's voice and the firm's rules, not "AI"; "is it built on Claude?" is answered truthfully, never led with (D23). Founding bands argue against pricing properly from the start (Superhuman's founder estimated holding launch pricing too long cost about 30% of ARR); we accept the trade, cap it at 1,000 seats and 31 December, and measure its cost. Failure modes to avoid: selling drafting alone, sprawl, faked traction (no logo without written consent), anti-human messaging, and scarcity that is not at a real price step.
The five numbers to watch (weekly)#
| # | Number | Planning hypothesis |
|---|---|---|
| 1 | First conversations held, how many met all six qualification tests, share ending with a dated next step | Next-step rate near 100% |
| 2 | Paid accounts and seats by band and vertical; self-serve wins separate | Base case about 75 seats by 31 December |
| 3 | 14-day activation | 40% of accounts per weekly cohort; counts until a cohort has 20 accounts |
| 4 | Month-one logo churn plus cancellation requests | 6% or less; pause promotion above 8% |
| 5 | Posting habit: paid seats posting at least 2 posts a week in at least 3 of the last 4 weeks, days 60 to 90 after payment | A majority of paid seats; counts under 20 seats; never punish a missed week |
Decisions needed now#
- D1 engagement data route and LinkedIn terms: display-only until Karl rules. The biggest risk in the plan.
- D15 with D18: founding terms in writing before the first payment.
- D4 Founding Design Partner terms.
- D3 with D29: launch motion and ready to take money (self-serve checkout with a person on hand; the site takes money at public launch; until D3 is decided, the 24 July NDA rule stands unless John waives it in writing).
- D22 launch video owner; Day 2 of launch week depends on it.
- D20 lawful basis for the 6,000-company list: nothing sent until Karl rules.
Before the founding close is announced in early December: D28, what a new small team is offered at Standard from 1 January.