Tactics
Every tactic with the hours it takes, the days until a signal, and how strong the evidence is.
31 tactics, newest first.
- Answer how-to questions in other products' forums
Community. Founders answer how-to questions in other products' forums, then mention their tool. Zapier's early customers came this way, at roughly 10 to 20 visitors per forum link.
CommunityEffort20 hTime to signal21 dEvidencePractitioner report - List the product in a platform's app marketplace
Partnerships. List your product in a platform's app marketplace. Standuply grew from about 200 to 1,000 teams after a Slack feature, and Importify saw day-one paid upgrades.
PartnershipsEffort60 hTime to signal30 dEvidenceCase with numbers - Pay writers for tutorials that answer your buyers' setup searches
Search. Pay outside writers for setup tutorials your buyers search for. DigitalOcean raised its typical payout to $300 per tutorial in 2018, and an outside estimate ties its library to 5M+ monthly visits.
SearchEffort40 hTime to signal90 dEvidencePractitioner report - Book customer interviews by cold email before you write code
Validation. Cold-email target customers for an hour-long interview before building anything. At WP Engine, the founder reports 40 of 50 agreed to talk and 30 became customers.
ValidationEffort25 hTime to signal14 dEvidenceCase with numbers - Make the premium tier the visual default
Pricing. Give the top pricing tier more space, larger type and the main button color. One third-party write-up reports 78% of Refactoring UI buyers chose it, alongside other changes.
PricingEffort12 hTime to signal28 dEvidencePractitioner report - Give the core product away and charge on transactions
Pricing. Make the core product free and earn on payments, payroll or loans run through it. Wave and Square show fast adoption, but their published numbers are reported, not tested.
PricingEffort80 hTime to signal45 dEvidenceCase with numbers - A free standalone micro-tool that feeds sign-ups to the main product
Free tool. A free, separately branded tool that does one narrow job and hands users into your main product. Eraser credits its DiagramGPT sidecar with 30% of new sign-ups.
Free toolEffort80 hTime to signal30 dEvidenceCase with numbers - A free tier for people who will never pay, to reach the people who will
Free tool. Give a capped free tier to people who will never pay but sit near buyers. Algolia and GreyNoise show large free usage, but neither ties those users to sales.
Free toolEffort60 hTime to signal90 dEvidenceCase with numbers - Judge freemium conversion by how long users have stayed, not by month one
Pricing. Track freemium conversion by signup cohort as it ages, not at month one. Evernote reported cohorts rising from 0.5% premium in month one to 5.5% after two years.
PricingEffort16 hTime to signal90 dEvidenceCase with numbers - Migrate switchers from a competitor by hand
Sales assist. Doing a prospect's migration off a competitor for them, free, after a good sales call. ConvertKit reported MRR up 54% to $3,237 while doing it, alongside other moves.
Sales assistEffort30 hTime to signal30 dEvidenceCase with numbers - List-style articles that answer buying searches and earn affiliate revenue
Search. Numbered best-of articles aimed at buying searches, with affiliate links. Urban Tastebud's founder says one list post started earning and the site now makes $51,000 a month.
SearchEffort30 hTime to signal90 dEvidenceCase with numbers - Skip the pre-launch and make setup one click so first users show it off
Activation. Cut setup to one click so new users see value at once and want to share it. Baremetrics skipped pre-launch tactics and reports 100 customers in about 4 months.
ActivationEffort60 hTime to signal30 dEvidenceCase with numbers - Publish your revenue numbers in public
Content. Post your real revenue on a fixed schedule as content. Groove and Baremetrics did this in public, but neither source ties signups or sales to the numbers.
ContentEffort20 hTime to signal60 dEvidenceCase with numbers - Open-source the tool, then sell the hosted version
Open source. Publish your tool's code for free, then charge for a hosted version. Sentry had a paying customer on day one at seven dollars. SimpleLogin shows open code used as a trust pitch.
Open sourceEffort80 hTime to signal30 dEvidenceCase with numbers - Own one niche keyword with organic pages and small search ads
Search. Pick one narrow keyword, build organic pages for it and buy small search ads. Bingo Card Creator got 56% of 2012 traffic from Google and 12% from AdWords, but credits sales gains elsewhere.
SearchEffort40 hTime to signal60 dEvidenceCase with numbers - Sell a paid product to the users of your free open-source project
Open source. Sell a paid add-on to users of your free open-source project. Tailwind UI made $400,000 on day one, and Forge reportedly had thousands of paying customers in a month.
Open sourceEffort120 hTime to signal30 dEvidenceCase with numbers - Pre-sell the product to your email list before you build it
Validation. Email your list a mock prototype and a paid pre-order before you build. Encharge got 50 pre-orders in two weeks from a list of about 500, against a goal of 5.
ValidationEffort20 hTime to signal14 dEvidenceCase with numbers - Pay for better photos of the supply side's listings
Marketplace. Pay photographers to shoot your suppliers' listings. Third-party write-ups report Airbnb's New York revenue doubled within a month, with no published baseline or control.
MarketplaceEffort40 hTime to signal30 dEvidencePractitioner report - Programmatic SEO: one page template, one page per long-tail query
Search. Programmatic SEO builds one templated page for each long-tail query. Avvo's directory reported 100M+ annual visits, but its sources never credit that traffic to the pages.
SearchEffort60 hTime to signal45 dEvidenceCase with numbers - Recruit a marketplace's supply side by hand first
Marketplace. Founders sign up a marketplace's first suppliers one at a time. Etsy worked craft shows, and Thumbtack emailed pros it found by hand, which a co-founder credits for its first 1,000 users.
MarketplaceEffort60 hTime to signal30 dEvidenceCase with numbers - Seed the users whose peers will see the product first
Product. Get the product to a small group whose peers watch them work. One newsletter says Granola spread through VCs, but it credits Granola's growth figure to design.
ProductEffort40 hTime to signal30 dEvidencePractitioner report - Send free product to small niche creators before launch
Creators. Mail free product to creators your buyers already follow, before sales open. Gymshark and Graza both did it with no marketing budget and report big launch spikes, self-reported and not isolated.
CreatorsEffort30 hTime to signal30 dEvidenceCase with numbers - A self-serve first purchase that grows into a company account
Product. A low-commitment self-serve purchase that later grows into a company plan. At Snappr, 48% of enterprise Workflows customers first made a self-serve Shoots booking.
ProductEffort80 hTime to signal90 dEvidenceCase with numbers - Make every link your customers share carry your product
Product. Put your product's name and a way to start on every link customers share. Gumroad's case shows fast early creator sales and traffic, from third-party write-ups only.
ProductEffort40 hTime to signal21 dEvidencePractitioner report - Launch on Hacker News
Launch. Post your working product on Hacker News yourself. Candy Japan reached 300 paying subscribers in four months; a third-party write-up reports PostHog had hundreds of users in a month.
LaunchEffort12 hTime to signal7 dEvidenceCase with numbers - Spin a habit your community already has out into its own product
Community. Move an activity your community already does informally into its own product and brand. 99designs did this with SitePoint design contests and later reported $15 million in annual revenue.
CommunityEffort120 hTime to signal30 dEvidenceCase with numbers - Let the support team, not sales reps, convert trial users
Sales assist. Support staff, not sales reps, help trial users onboard and convert. Pipedrive's co-founder credits this model, but the only figures are a self-reported early customer count.
Sales assistEffort40 hTime to signal45 dEvidenceCase with numbers - Trade swag for a product install
Activation. Offer free swag only to users who finish your install step. One vendor-reported case, New Relic's Data Nerd shirt, claims a 30% rise in qualified leads, meaning agent installs.
ActivationEffort30 hTime to signal21 dEvidencePractitioner report - Buy the typo of your domain and catch the lost sign-ups
Conversion. Buy the misspelled or wrong-ending versions of your domain and redirect them. Testimonial.to's founder paid $35,000 for the .io and says it brought in $80K in lifetime revenue.
ConversionEffort10 hTime to signal30 dEvidenceCase with numbers - Let people use the product before they sign up
Activation. Let visitors use the real product before asking for an account. Codecademy and Rows both did it, and Rows reports 72% more sign-ups, but neither case isolates the effect.
ActivationEffort40 hTime to signal14 dEvidenceCase with numbers - Video-style thumbnails in marketing emails
Email. Overlay a play button and time stamp on an email image so it reads as a video. Wistia's seven A/B tests averaged a 21.52% click lift, but the gain shrank in later sends.
EmailEffort6 hTime to signal3 dEvidenceCase with numbers
Tools for these pages
Calculators and templates these pages link to, each with its formula or method on the page.
- A/B test sample size and duration calculator, with the formula
A 5% baseline needs 31,234 visitors per variant to detect a 10% relative lift at 5% significance and 80% power. Formula, inputs and run time.
- Growth experiment template with ICE scoring, a weekly cadence and a backlog you can copy
A growth experiment template gives each idea a hypothesis, an ICE score from 1 to 10, one owner and one metric, then ranks the backlog every week.
- Growth model template for startups, with inputs, a cohort layer cake, a loop row and a sensitivity table
A growth model template in four sheets: inputs, a cohort layer cake, a loop row and a sensitivity table, with worked numbers you can copy this week.
- Onboarding Email Sequence Template With Six Emails, Timing and Metrics
Six onboarding emails from Day 0 to Day 21+, each with a trigger, one job, an exit condition and one metric, measured against a holdout group.
- Viral Coefficient Calculator With the K Factor Formula and Viral Cycle Time
The viral coefficient (K) equals invites per user times invite conversion rate. Above 1 it compounds. Below 1, total users equal seed divided by (1 minus K).
Elsewhere on Race to Scale
- Guides
How the mechanics work, with the evidence and the arithmetic.
5 pages - Benchmarks
Split times for growth metrics, each with its source and its sample.
5 pages - Teardowns
Real pages redrawn as their structure, with what each part does.
3 pages - Tools
Calculators and templates, with the formula and a worked example on the page.
7 pages - Playbooks
Several tactics in order, with the timings between them.
7 pages - Cases
Short cited accounts of what a company ran and what happened.
43 pages - Digests
Public reports read for you, with their key numbers.
1 page - Glossary
Growth terms defined in a sentence or two, with where they apply.
12 pages - Agencies
Growth agencies for startups and small companies, A to Z, with facts checked on each agency's own pages.
- Voices
Posts from people who run growth that teach something specific, with our label and our take.
How these pages are made
Every factual sentence on a Race to Scale page is checked against a passage on a source page that was read for that page, and numbers are computed in code from those sources, not typed from memory. Each page lists its sources and links them where they are used.
Every page shows the day it was first published and, when a fact changes, the day it was updated; no date is set earlier than the day a page went up. The full method is on How we work.