Seven Companies, Seven Ways to Acquire Customers Without Ads

Why these seven

Seven companies, each running a distinctly different acquisition motion. The point is not the company list, it is the seven mechanics underneath.

Everything below is drawn from public evidence: funding announcements, published comparison pages, open source repositories, press coverage, and the companies' own sites. We have no inside knowledge of their internal numbers, and where a figure is reported rather than confirmed we say so.

One thing is worth stating before the detail. Across all seven, paid media is essentially invisible. Growth is community, content, comparison, referral, retail, and press. That is not a coincidence and it is the most useful finding here.

Company	Category	Core motion
Lassie	AI back office for medical and dental	Community embedding, referral compounding
Rerun	Robotics data infrastructure	Open source funnel into a commercial tier
Griffin	UK banking as a service	Regulatory milestones as marketing events
Dix Hectares	Luxury skincare	Provenance narrative plus designed repurchase
Norma	Physical focus device	Comparison led capture, anti subscription
Supercut	Async video messaging	Competitor displacement
Bobbin	AI tutoring platform	Borrowed authority, gated waitlist

1. Lassie: become a participant, not a vendor

AI agents running back office operations for independent medical and dental practices. Founded by Steijn Pelle and Frederic Renken, previously product leaders at Robinhood, Coinbase and Superhuman. Reportedly over 700 practices, and a $35M Series A led by a16z in June 2026.

They did not run acquisition campaigns. The founders embedded themselves in the customer's world, going as far as working as dental practice administrators before building the product. Discovery ran through dentist study groups, notably Invisalign study clubs. Growth from there has been overwhelmingly referral driven.

Why it works. Dentistry is a high trust, high density referral network where practice owners talk constantly. One credible reference travels further than any advertisement. And the proof point is checkable in ten seconds: a practice loses 100 or more hours a month to admin, against roughly $200K a year in staffing for it. Lassie claims an average saving of at least 30 hours. A practice owner can test that against their own payroll immediately, which is the specificity that survives a sceptic.

The mechanic. In a fragmented vertical with a dense professional network, the cheapest channel is going where buyers already gather and becoming a participant rather than a vendor. Slow to start, then compounds without spend, exactly like referral systems generally.

The limit. This is why they raised. Referrals do not take you from 700 practices to 7,000, and they now have to build a scalable channel from a standing start.

2. Rerun: give away the habit, sell the scale

Stockholm based, building visualization and data infrastructure for robotics and computer vision. Open source SDK and viewer, free under permissive dual licensing chosen so the tool can be adopted individually and absorbed into enterprise stacks without legal friction. The commercial product is a data catalog sold to teams.

Why it works. The free tier is genuinely load bearing: developers at Meta, Google and Hugging Face use it. That is not a purchased logo wall, it is a byproduct of the tool being good, and it is far stronger proof than any customer quote. Roughly 8,000 GitHub stars and an active Discord do support, evangelism and product feedback at close to zero marginal cost.

The upgrade path is natural rather than coercive: use it free locally, pay when your data outgrows your machine. Nobody has to switch tools or relearn anything, which removes the usual reason people refuse to convert.

The mechanic. Split the product along a usage boundary the customer will cross on their own as they succeed. Free for the individual, paid for the team. The upgrade trigger is the customer's growth, not a feature you withheld.

3. Griffin: publish the journey, not just the destination

UK bank and banking as a service platform. Applied for a licence in May 2022, entered mobilisation March 2023, fully authorised March 2024.

In a category where trust is the entire product, Griffin turned the regulatory process itself into a content engine. Each stage was published as an event. They pair it with transparency as positioning: a public company facts page listing their FCA firm reference number, Companies House link, full licence timeline, and investor list.

Why it works. Getting a UK banking licence is genuinely hard, and doing it in roughly a year is a story journalists write for free. Transparency here is a costly signal rather than a compliance chore: publishing your regulator reference number is something competitors operating on someone else's licence cannot copy.

They also hired a marketer from developer focused SaaS rather than banking, which shows in the output: free sandbox, clear docs, transparent pricing. They sell to engineers the way developer tools companies do, in a category where everyone else sells to executives.

The mechanic. If your industry has slow, painful, verifiable milestones such as licensing, certification or audit, publish the journey. It generates a content calendar out of work you are already doing, and the difficulty is the moat.

4. Dix Hectares: engineer the repurchase into the product

French luxury skincare, cultivating botanical actives since 2021 on a ten hectare plot on the Château Montrose estate. Three reinforcing layers.

Provenance. Built on the terroir of a Grand Cru wine estate. Hand harvested at peak potency, fresh frozen within six hours, low temperature extraction, full traceability. In French luxury this is the entire credibility structure, and it is unrepeatable because it depends on owning that specific land.

Proprietary science. They trademarked their own methodology, which means comparisons happen on their terms.

Seasonal cadence. This is the commercially clever part. Products are released and reformulated by season, sized to last about three months. The product cycle and the purchase cycle are the same cycle. Repurchase is not a retention campaign, it is designed into the range.

They also launched with only two products despite pressure to do more. That restraint is itself an asset in a category defined by overwhelming SKU counts.

The mechanic. Find the one thing about your operation a competitor physically cannot replicate and build the brand on it rather than on features. Where possible, engineer the repurchase interval into the product format instead of chasing it with lifecycle email.

5. Norma: write the honest comparison

A 70mm machined stainless steel NFC disc. Scan it and your chosen apps are blocked until you scan again. No battery, no subscription, bought once.

Their most substantial published asset is a direct, named comparison against their two competitors. Crucially it is argued rather than asserted. It concedes that one rival is powerful software, then explains that its off switch lives on the same phone you are escaping and the blocking stops when you stop paying. It grants the other real credit as a physical, buy once device, then notes it is plastic and single platform.

Why it works. People searching for a screen time solution search competitor names, not generic terms. Owning that comparison intercepts high intent traffic at the moment of evaluation. And the honesty does the persuading: a page that trashes the alternatives reads as advertising, a page that concedes their strengths reads as advice.

Anti subscription positioning converts a business model constraint into a brand promise, and placement in curated design retail reaches a higher intent buyer than the digital wellbeing crowd.

The mechanic. Write the honest comparison page against your named competitors. Concede what they are genuinely better at. It captures the highest intent traffic in any category and costs one piece of content.

The gap. They sell a tactile object almost entirely in text. The weight, the machining, the physical reach for the disc, none of it is carried by the marketing.

6. Supercut: name the reason people are unhappy

AI powered async video messaging with automatic editing. Founded by people who worked together at Typeform Labs.

They are not competing for a new category. They target a specific dissatisfied user base: people who loved an incumbent and feel it declined after acquisition. Reviewers repeat that framing almost verbatim, which means the narrative has propagated successfully. They also position against the opposite failure mode, tools that are too heavy and require real editing time. One approach trades quality, the other trades speed, and Supercut claims the middle.

Their organic presence is dominated by other people's platforms rather than owned content, because independent evaluations carry more weight for a tool purchase. And the free tier gives unlimited viewer seats, which is structurally correct: the buying decision is made by the recipients, not the recorder.

The mechanic. Find an incumbent whose users are unhappy for a specific, articulable reason, then name that reason. Displacement is far cheaper than category creation, because the audience already understands the product and already has budget. Then structure your free tier around who actually decides, which is often not the person who signs up.

7. Bobbin: launch into an audience you already have

AI lesson planning and parent reporting for tutors, built by a London company with over a decade in tutoring business software, and launched inside that parent brand rather than as a new name.

The parent runs a substantial blog ranking for the queries their buyers search. Bobbin gets mentioned inside content the audience was already reading for other reasons, which is the cheapest possible product introduction because the traffic already exists. Access is deliberately gated to existing customers plus an early access register, which controls load, manufactures scarcity, and builds a warm list for the standalone launch.

The mechanic. If you have any existing audience, launch new products into it before launching them to the market. Gate early access deliberately, because the list you build during the gate is the launch.

The pending risk. The strategy is a loan. Going standalone means facing a different buyer without the parent's credibility attached.

The four patterns that repeat

1. Own something that cannot be copied, then build everything on it. Dix Hectares has land. Griffin has a licence. Rerun has genuine open source adoption. Lassie has embedded operational knowledge. In every case the marketing points at an asset a competitor would need years or millions to match. Feature marketing is a race. Asset marketing is a moat, and it is the practical version of category entry points.

2. Give away the thing that creates the habit. Rerun gives away the daily developer tool. Griffin gives a free sandbox. Supercut gives unlimited viewer seats. The free layer is where the habit forms and the internal advocate is created. The paid layer sits where the customer's own success makes it necessary.

3. Concede to your competitors, in writing. Norma's comparison page grants a rival real credit before explaining where it stops. Supercut acknowledges the incumbent was excellent before it declined. Naming a competitor's genuine strengths increases conversion, because it signals you are describing reality rather than selling.

4. Design the repurchase rather than chasing it. Dix Hectares sizes formulations to run out at the season's end. Where the product format can carry retention, that is far cheaper than any campaign, and it beats the retention maths most businesses fight.

What is missing across all seven

The absences are as instructive as the plays.

Almost no paid acquisition. Worth sitting with, given how many small companies treat ads as the default first move.

Very little video. Only Supercut, whose product is video, has meaningful video presence. Norma sells a tactile object through text. Dix Hectares sells sensory luxury with a heavily written site. Bobbin sells to an audience that lives on YouTube with a blog. This is the single largest consistent gap in the set, and it aligns with where attention actually is.

No creator programs, including in the two consumer brands where the category norm expects them.

Founder history is doing real marketing work. Ex Robinhood, ex Typeform, ex CircleCI, the creator of a widely used open source framework. In small companies the founders' background is often the reason press coverage exists at all, which is the argument for founder led content.

Frequently asked

Does this mean ads do not work? No. It means these seven built compounding channels first, and none of them needed to buy demand to reach this stage. Paid amplifies a funnel that already converts.

Which mechanic should we copy? The one matching your structure. Dense professional network: embed. Developer audience: open source. Regulated category: publish the journey. Consumable product: design the repurchase. Crowded category with a disliked incumbent: displacement.

Can Fixora work out which applies to us? That is the first thing we do. Send us your market and get the read within 48 hours.