What thirty small-business clients actually ask for
Seven years of inbound requests, grouped. The pattern decided what we build, and it was not what we would have guessed.
We maintain thirty live client sites. Almost all are US small businesses, built and supported in-house since 2019. No sales team, no outbound — the whole book came from referrals compounding.
When we grouped them by what the business actually does, the shape was not what we expected:
- Home and trade services — 5
- Healthcare, home care and staffing — 5
- Nonprofit, faith and community organisations — 5
- Transport and logistics — 3
- Real estate, insurance and finance — 3
- Retail, media and other — 9
The requests do not match the build work
Here is the thing that reframed the business for us. What clients pay us to build is websites. What they ask us for, unprompted and repeatedly, is marketing.
Roughly in order of frequency:
- “Can you post something for us?” Blog posts, service pages, seasonal updates. We write these for most of the active clients.
- “Can you run our ads?” Meta mostly, some Google. We manage campaigns for a subset.
- “Can you make us a video?” This is the one we have been least able to say yes to without absorbing a cost.
Nobody asks for a website redesign. They ask for customers.
Where video demand concentrates
Two clusters, for completely different reasons.
Home and trade services need offer-led creative that goes stale on a season cycle. The AC repair ad that works in July is dead in October. Every rotation needs new creative, and the offer changes faster than any shoot schedule could keep up with.
Home care and healthcare staffing need recruiting creative, and this was the surprise. Their ad spend is not aimed at customers, it is aimed at hiring caregivers and nurses. Recruiting creative burns out much faster than customer creative because the audience is smaller and sees it repeatedly. These clients need something new every few weeks.
Between them that is ten of thirty clients with recurring, non-optional video demand. Not a segment we picked. A segment that has been sending us requests.
The nonprofit cluster we have not solved
Five community and faith organisations, including a Haitian community association, a Guyanese cultural site and a Yoruba heritage organisation.
They ask for video too — event promos, appeals, announcements. But their budgets are close to zero and some of their audiences are not English-first. We do not have a good answer for them yet and we are not going to pretend the product serves them on day one. It is on the list because they asked, not because we have figured it out.
What this decided
We were close to building something entirely different — a conversational intake product, sitting on the healthcare staffing cluster. It was a defensible idea and it would have been a real product.
It also was not what anyone had asked us for.
Thirty clients, seven years of inbound, and the most frequent unmet request is video for paid campaigns. Building the thing being asked for is a lower-variance bet than building the thing we found interesting, and after seven years of slow growth we would rather take the lower-variance bet.