# The first growth machine is usually hand-built > Early growth often starts with awkward manual work: filling blank states, writing the useful post, narrowing the list, and using real signals before you automate anything. - Canonical HTML: https://growth.iangoh.com/blog/the-first-growth-machine-is-usually-hand-built/ - Published: 2026-05-24 - Updated: 2026-05-24 - Categories: early traction, operator-led distribution - Niches: SaaS, creator tools, SEO, outbound ## On this page - Blank screens kill more demand than ugly workflows - The useful post can carry the commercial pages - Make the free experience do two jobs - A tiny target list makes better founders - Where this matters ## Start with these related tactics - [Manual empty-state concierge onboarding](/growth-ideas/manual-empty-state-concierge-onboarding/): For the first few dozen users, fill the blank workspace or report yourself so the product starts with proof instead of an empty screen. - [Startup-learning post backlink wedge](/growth-ideas/startup-learning-post-backlink-wedge/): Publish one concrete builder story on the product domain, then let that honest post earn links that lift the rest of the site. - [Watermarked no-signup free tier loop](/growth-ideas/watermarked-no-signup-free-tier-loop/): Let free users get the core outcome without creating an account, but keep a visible watermark so distribution and upgrade pressure happen in the same flow. Founders like to talk about building a growth machine. Fair enough. The trouble is that the first useful version rarely looks like a machine. It usually looks like somebody doing the work by hand. A founder fills the first empty account so the product does not feel dead. Someone writes the one honest post that earns links while the product page cannot. The outbound list gets cut from thousands to twenty because twenty can still be treated like people. ## Blank screens kill more demand than ugly workflows Buffer Analyze is a good reminder that activation is often embarrassingly manual before it becomes elegant. Tom Redman described a [manual empty-state concierge onboarding](/growth-ideas/manual-empty-state-concierge-onboarding/) phase where the team created reports on users' behalf so they would not land in an empty product. That is not a hack in the pejorative sense. It is simply refusing to make the user do all the imagination. A blank state asks for faith. A populated one asks for a judgment. Early growth gets easier when the user can react to something real. ## The useful post can carry the commercial pages Kapwing's early story is still one of my favorite SEO lessons because it did not begin with some grand topical authority plan. It began with a sharp builder writeup. The [startup-learning post backlink wedge](/growth-ideas/startup-learning-post-backlink-wedge/) worked because people were willing to link to a specific lesson about the product's payment flow. That is the part teams often miss. The page that earns links and the page that converts do not have to be the same page. One can do the carrying for the other if it lives on the same domain and says something worth citing. ## Make the free experience do two jobs VEED had very little runway and no reason to wait for a perfect funnel. What it had was a product people could use quickly. The [watermarked no-signup free tier loop](/growth-ideas/watermarked-no-signup-free-tier-loop/) let users get the result before they created an account, while the watermark turned free usage into distribution and the removal upgrade turned delight into revenue. I like this because it forces honesty about what the product is for. If the free use is satisfying enough to spread but incomplete enough to upgrade from, the product is already teaching you where the line belongs. ## A tiny target list makes better founders The same logic shows up in outbound. Chris Bakke's [20-account creative prospecting sprint](/growth-ideas/twenty-account-creative-prospecting-sprint/) is really an argument against fake scale. A list of 7,000 leads makes it easy to hide behind tooling. A list of 20 forces you to ask what would actually get a reply from this company. That is also why the [job-post signal email opener](/growth-ideas/job-post-signal-email-opener/) is useful. It starts with something current and verifiable instead of pretending every prospect woke up hoping for your pitch. One company is hiring. One role is open. One sentence from the job spec gives you the bridge. ## Where this matters For SaaS, build the first onboarding by hand if the alternative is an empty app and a lost user. For creator tools, let the free output advertise the product while teaching users why the paid version exists. For SEO-heavy products, publish the builder story that people actually want to cite. For outbound, trade fake personalization at scale for a much smaller list with much better reasons to care. The mistake is thinking manual work is the opposite of a system. Often it is the draft of the system. If a hand-built version keeps producing signups, replies, upgrades, or links, then you have earned the right to automate it. Before that, automation is usually just a faster way to stay vague. ## Related GrowthDex tactics - [Manual empty-state concierge onboarding](/growth-ideas/manual-empty-state-concierge-onboarding/) - Product, Email, Onboarding - [Startup-learning post backlink wedge](/growth-ideas/startup-learning-post-backlink-wedge/) - SEO, Content, Hacker News - [Watermarked no-signup free tier loop](/growth-ideas/watermarked-no-signup-free-tier-loop/) - Product, Referrals, Website - [20-account creative prospecting sprint](/growth-ideas/twenty-account-creative-prospecting-sprint/) - Outbound, LinkedIn, Content - [Job-post signal email opener](/growth-ideas/job-post-signal-email-opener/) - Email, Outbound, Sales ## Essay chronology - [Newer essay: The first customers are supposed to feel a little unfair](/blog/first-customers-are-supposed-to-feel-a-little-unfair/) - early traction, operator-led distribution - [Older essay: The launch keeps going after launch day](/blog/the-launch-keeps-going-after-launch-day/) - launch strategy, brand trust ## Keep reading - [Compounding growth usually waits for density](/blog/compounding-growth-usually-waits-for-density/) - SEO, operator-led distribution - [The buyer is usually already on the right page](/blog/the-buyer-is-usually-already-on-the-right-page/) - SEO, conversion, operator-led distribution - [The first customers are supposed to feel a little unfair](/blog/first-customers-are-supposed-to-feel-a-little-unfair/) - early traction, operator-led distribution ## Continue through the blog - [SaaS](/blog/#path-saas) - 3 essays in this path ## Sources - [Buffer Open Blog](https://buffer.com/resources/buffer-analyze-journey/) · [GrowthDex source hub](/sources/buffer-open-blog-buffer-com/) - [Kapwing Company Blog](https://www.kapwing.com/blog/how-we-got-our-first-10-customers/) · [GrowthDex source hub](/sources/kapwing-company-blog-kapwing-com/) - [VEED Learn](https://www.veed.io/learn/how-to-get-first-paid-users-saas) · [GrowthDex source hub](/sources/veed-learn-veed-io/) - [Product Hunt Stories](https://www.producthunt.com/stories/a-better-way-to-get-your-first-10-b2b-customers) · [GrowthDex source hub](/sources/product-hunt-stories-producthunt-com/) - [Product Hunt Stories](https://www.producthunt.com/stories/the-cold-email-template-that-got-me-an-8-reply-rate) · [GrowthDex source hub](/sources/product-hunt-stories-producthunt-com/) ## Editing notes - Kept the essay as one argument about manual work turning into systems instead of a tactic roundup. - Cut generic startup inspiration language and stayed close to the concrete cases and numbers. - Used short paragraphs, direct opinions, and plain verbs to avoid polished AI cadence. - Made the internal links carry the reasoning so they help readers and crawl depth at the same time. ## Advisory If you want help turning this into a growth system, Ian Goh offers advisory at https://iangoh.com/advisory.