Funnel advice fails small businesses in a specific way: it is written by people selling funnel software, for companies with funnel teams. The seven-stage diagram is not wrong. It is just irrelevant when the whole marketing department is the owner, on Thursday evenings.
Here is the version that works at small scale. It has three parts.
One offer, stated like a human
Most small business marketing fails before any funnel exists, at the sentence level. If you cannot say who the offer is for, what changes for them, and what it costs to find out, in three sentences a stranger understands, no funnel will save the campaign built on top.
The test I use: read your homepage sentence to someone outside your industry. If their next question is "so what do you actually do", stop building funnels and fix the sentence.
One route, fully built
Pick a single route from stranger to customer and build every step of it before touching a second one.
A route means: one channel where your buyers already are, one piece of the offer they can say yes to cheaply, and one follow-up path that does not depend on them remembering you. For a service business that might be: local search, a fifteen-minute consultation, and a reminder sequence. For a product: paid social, the product page, an abandoned-checkout flow.
The discipline is in the word one. Two half-built routes convert worse than one finished route, every time, and small businesses almost always run four half-built routes because every channel guru insisted theirs was mandatory.
One number, watched weekly
Not fifteen dashboards. One number that summarises whether the route works: cost per enquiry, or cost per first order. Look at it weekly, with last week beside it. Everything else, opens, likes, impressions, reach, is diagnosis for when the one number moves the wrong way, not a scoreboard to celebrate.
When the number is stable and acceptable, you have earned the right to add a second route. Not before.
Where AI fits, and where it does not
AI collapses the production cost of a funnel: pages, ads, variations, follow-up emails, all faster than a year ago. What it does not collapse is the two decisions that make the funnel work, the offer and the route. Produce with the machine. Decide like an owner.
I am writing my second book on exactly this territory, marketing for small businesses, because the gap between what small companies need and what marketing content gives them is where I have spent much of my working life. My first book, Winning the Slow Race, is about building the company itself; the funnel above is the marketing chapter of how I think.
If you want this applied to your specific business rather than in general, that is a working session, and it is usually one session, not a retainer.