Lean GTM Plays

Ten motions, each grounded in named public evidence — a founder's own blog post, a documented case, one repeatedly-cited growth book — compressed for a team that doesn't have a growth department, just someone who can build a sheet and send an email. Evidence below is cited as technique, never as a promised result: nobody's founder-led cold email is your founder-led cold email. Hour estimates are a rough builder's floor, not measured data — treat them as a planning number, not a promise.

Each entry: the symptom that calls for the play, the smallest version worth building under roughly 30-40 people, what it costs in hours, and the number that tells you it failed.

Replies are the only real signal you have before your first ten customers.

The 1:1 Outbound Sprint · tags: pre-pmf · founder-led · outbound — evidence anchor: Proxycurl's founder started with cold email to the first 10 customers; Churnkey closed its first 10 via founder email and chat in about three weeks (both self-reported, cite as technique).

Symptom

pre-PMF, a handful of customers or fewer, no idea who actually replies to what.

Smallest version

one sheet, three tabs — accounts, messages, replies. 30–50 hand-picked accounts, one named sender, 5–10 tailored emails a day. No CRM, no automation.

Cost

~6–10 hours to build the sheet and load the first 50 accounts. Sending is a daily habit after that, not a build cost.

Failure signal

fewer than a handful of real replies after 50 sends, or zero repeated objection language across the replies you do get.

Paid spend before a shared definition of "qualified" is a guess with an invoice.

The Channel Staircase · tags: channels · cac · test-log

Symptom

arguments about which channel to try next, nobody able to say what a lead costs by channel.

Smallest version

one sheet — backlog, test log, dashboard, config — tracking 2–4 channels against one conversion definition the whole team agreed to on a single call.

Cost

~5–8 hours to build. The tests themselves run on their own clock, 2–4 weeks each.

Failure signal

30 days in, no test has a written keep, kill, or scale decision. An undecided test is a stalled staircase, not a slow one.

A founder's best material dies in the call it was said in.

The Founder Megaphone · tags: founder-led · content · distribution — evidence anchor: Lovable ran founder-led distribution at lean headcount with no separate marketing org (cite as inspiration, not a promised replica).

Symptom

a founder who says something worth publishing on every customer call, and posts about twice a quarter.

Smallest version

one sheet — idea inbox, production board, tracker — fed by whatever the founder already uses to capture thoughts, voice memo or Slack DM. No automation required at this size.

Cost

~4–6 hours to build the sheet. The founder's writing time is the real cost: budget 2–3 hours a week, on a calendar, not "when there's time."

Failure signal

the idea inbox fills and the production board doesn't move for two weeks straight. A full inbox with nothing posted is a queue, not a megaphone.

A free tier that can't name its own activation event can't tell you who's ready to buy.

The Freemium Wedge · tags: freemium · activation · scoring

Symptom

a live free tier or trial with real signups, and nobody able to say which behavior predicts an upgrade.

Smallest version

one tracked activation event, one sheet scoring accounts against it, a Slack alert when someone crosses the line. Skip lifecycle email until the scoring is proven.

Cost

~8–12 hours, assuming the event is already trackable. If it isn't, that's an engineering task first, not a GTM one — don't start the sheet until it is.

Failure signal

flagged accounts upgrade at roughly the same rate as unflagged ones. The score isn't finding anything the whole free base wouldn't already tell you.

A launch machine outlives the launch. A launch team does not.

The Launch-Day Blitz · tags: launch · community · timing — evidence anchor: EarlyBird's coordinated same-morning launch across Product Hunt, HN, Reddit, and community channels (self-reported, cite as pattern).

Symptom

a dated launch, three to eight weeks out, and a real fear that launch morning turns into chaos.

Smallest version

one sheet — checklist, calendar, dashboard, queue — a T-minus-14 schedule, one anchor channel where the founder already has standing, two named humans on launch morning.

Cost

~6–10 hours to build and dry-run once, a week before launch.

Failure signal

planned posts go out more than 30 minutes late, or a reply sits unclaimed for hours. The whole machine's job is timing — a late machine failed at its one job.

A listing with no reviews and no route back to your CRM converts nobody.

The Marketplace Piggyback · tags: marketplaces · listings · reviews

Symptom

an integration that exists or is close, and buyers who already browse a specific marketplace or directory.

Smallest version

the listing itself, plus one sheet routing every install or lead into whatever you already use as a CRM, tagged by source. Ask for reviews manually at first.

Cost

~5–8 hours for the routing sheet. The listing approval clock belongs to the platform, not you, and can run weeks on its own.

Failure signal

installs climb but the CRM shows zero marketplace-tagged leads. The routing broke, or it was never really wired.

One template times a thousand empty rows is spam, not an engine.

The Page Generation Engine · tags: seo · programmatic-pages · integrations — evidence anchor: Zapier's integration-page network, one template plus a spreadsheet of variables (cite as technique, not a promised traffic number).

Symptom

a real axis of variation — integrations, industries, use cases — that search data shows buyers actually query, with no page for any of it.

Smallest version

10–25 pages, one pattern, published manually or by CSV import, not an API pipeline. At least two fields of genuine per-page proof, a screenshot or a real example, so it isn't thin content wearing a template.

Cost

~10–15 hours for the first small batch, most of it gathering proof, not building the mechanism.

Failure signal

fewer than half the batch is indexed 30 days after publishing. That's the signal to stop publishing and fix quality, not to publish more.

A verbal yes is worth exactly as much as a maybe.

The Pre-Sale Validator · tags: validation · pre-sale · pricing — evidence anchor: Subscribr pre-sold 50 lifetime seats with a firm delivery deadline and a full refund guarantee (cite as technique, not a promised outcome).

Symptom

an offer you could describe but haven't built, and no appetite to spend a quarter finding out nobody wants it.

Smallest version

one offer page, a real price, a written refund promise, a payment link, a one-sheet scorecard per call, a hard threshold — say, 10 payments in 4 weeks — written down before the first call happens.

Cost

~4–6 hours to build the page and tracker. The real cost is 10–20 founder-run sales calls.

Failure signal

the threshold isn't hit by the deadline. That's not a failed test, it's the test working. Building anyway past that point is the actual failure.

A referral loop multiplies existing love. It cannot create it.

The Two-Sided Referral Loop · tags: referral · retention · lifecycle-email — evidence anchor: Dropbox's two-sided referral drove growth from 100K to 4M users in 15 months (2009-era evidence; referral virality is widely believed weaker today — cite the mechanic, not the multiplier).

Symptom

an active, retained user base with no organized way to turn happy users into new ones.

Smallest version

unique referral links, a reward gated on the referred user's activation, not their signup, one sheet tracking invited, accepted, activated, retained. Invites go out by lifecycle email, not a product change.

Cost

~8–12 hours, assuming a lifecycle email tool already exists. Add real time if the invite needs a product surface built first.

Failure signal

referred users retain worse than everyone else. That means the reward recruited the wrong people, not that the loop needs more volume.

A waitlist only ranks demand if some people rank below others.

The Waitlist Ladder · tags: waitlist · qualification · launch — evidence anchor: Robinhood's gamified referral waitlist passed 1M; Superhuman ran 180K with a mandatory screening call (widely reported; cite as a pattern, not as a number to plan against).

Symptom

a launch or feature one to four months out, one sharp audience, and a plan to just collect emails.

Smallest version

a form with 2–3 qualification questions, a sheet scoring and segmenting signups into early user, watcher, or poor fit, manual weekly invites from the top of the list.

Cost

~5–8 hours to build the form, sheet, and scoring.

Failure signal

everyone who signs up scores about the same. A ladder with no bottom rung is a mailing list wearing a costume.

Sheet: tab plays, one row per play you actually started.

play,tags,started,smallest_version,hours_est,failure_signal,status,decided_on

The 1:1 Outbound Sprint,"pre-pmf · founder-led · outbound",2026-09-06,"one sheet, three tabs, 50 accounts, one sender, 5–10 emails a day",8,"fewer than a handful of real replies after 50 sends",running,

status ∈ {running, kept, killed}. decided_on is the date the failure signal was read and a call was made — an empty cell past the play's own clock is the stall the Channel Staircase warns about.