A marketing strategy is one decision, made in advance: where your money and effort go, in what order, and what you’re saying no to. Everything else is execution.
Most small businesses don’t have one. They have a website someone built, a social account someone runs, and a history of saying yes to whoever called last. A strategy replaces that with a short written plan you chose on purpose. If you’ve been following this series, you’ve already done the hard part, and AI can now assemble it in an hour.
Why this matters
Marketing without a strategy isn’t neutral. It has a predictable failure pattern, and it’s expensive.
- Scattered spending buys nothing that compounds. Three months of ads here, a burst of posting there. Each effort resets to zero because none of them was given time or a job.
- Without a plan, salespeople make your strategy for you. Every directory, ad platform, and agency that calls has a pitch. If you can’t check the offer against a written plan, the most persuasive caller wins, and you fund their strategy instead of yours.
- Channel-hopping feels like work but measures like noise. You can be busy with marketing every week and still have no idea what’s producing customers. Motion isn’t progress; sequence is.
- The order matters as much as the choices. Ads pointed at a weak website waste money. Reviews matter before local search pays off. A strategy sequences the work so each piece supports the next.
Saying no is the real product here. A one-page strategy gives you grounds to decline the sponsorship, the trend, and the pitch, without wondering if you’re missing out.
What a small business strategy contains
Skip the 40-page deck. Five short sections cover a business your size.
1. The channels you chose.
Which three or four places you’ll show up, and one honest sentence on why each fits your customer. Not every channel. Yours. For most local service businesses that shortlist starts with your website, search, reviews, and one social platform where your customers actually spend time.
2. The phases.
What gets fixed first, second, and third. A common sequence: make the website able to convert visitors, then work on getting found, then add paid reach once the foundation converts. Phases stop you from paying to send traffic to a page that loses it.
3. The budget.
A number per month, split across the channels, with a floor of patience attached. Search work is a six-month commitment. Ads can be tested in one. Writing the timeline down protects you from quitting the slow channels right before they pay.
4. The not-doing list.
The channels you’re deliberately skipping and why. This is the section that saves you money all year.
5. The numbers that mean it’s working.
Two or three real measures, like calls, quote requests, and booked jobs. Pick them now, before any vendor picks friendlier ones for you.
Putting it to work
The strategy is only as good as what it’s built from. That’s why it comes sixth in this series, not first.
- Gather your documents. Your industry analysis, your positioning and customer profiles, and your voice of the customer findings do the heavy lifting. Your brand style guide and voice guide shape how everything sounds and looks. Missing some? The prompt works with what you have; it’s honest about being weaker without the research.
- Tell the AI your real constraints. Monthly budget, hours you can give it, and how soon you need results. A strategy that ignores your capacity is fiction with headings.
- Run the prompt and argue with the draft. You know things the AI doesn’t: which jobs are profitable, what season carries you, what you’ll actually keep doing. Push back and make it revise. The second version is usually the keeper.
- Check it with money math. For each channel, ask what a customer is worth and what reaching one costs. If the AI can’t show the path from spend to customer in plain arithmetic, cut that channel.
- Write it on one page and date it. Then live with it for a quarter before changing anything. A strategy you revise monthly isn’t a strategy; it’s channel-hopping with paperwork.
One honest limit. AI assembles a solid strategy from good inputs, but it can’t know your market’s mood, your capacity in busy season, or which customers you enjoy. Treat the draft as a strong proposal from a well-read advisor. You’re the one who has to fund it and live it, so the final call on every line stays with you.
Next in the series: the metrics worth tracking, so you’ll know whether the plan is working without drowning in dashboards.
Take this prompt with you
Copy this into ChatGPT, Claude, or Gemini, and paste in whichever series documents you have.
Act as a marketing strategist for a small business. Build me a one-page marketing strategy from the materials below.
My business: [what you do, where, and roughly what a typical customer is worth to you]
My constraints: [monthly marketing budget] / [hours per week I can give marketing] / [when I need results: e.g. "steady leads by spring"]
My research documents, pasted below: [paste your industry analysis, positioning and customer profiles, and voice-of-the-customer findings. Paste what you have; skip what you don't.]
Produce exactly five sections:
1. Channels: the 3 or 4 channels I should be in, each with one sentence on why it fits my customer, drawn from my research.
2. Phases: what to fix or build first, second, and third, with a rough month range for each phase. Sequence it so nothing gets traffic before it can convert that traffic.
3. Budget: split my stated budget across the channels by phase, and tell me the minimum number of months each channel needs before I'm allowed to judge it.
4. Not doing: 3 to 5 channels or tactics I should deliberately skip for now, each with a one-line reason. Be specific enough that I can quote it when a salesperson calls.
5. Success numbers: 2 or 3 plain measures that tell me it's working, tied to customers and revenue, not clicks or followers.
Rules: stay inside my budget and hours. No channel gets chosen without a reason traceable to my research or my constraints. If my documents are missing or thin, list your top 3 assumptions and mark them clearly so I know what to verify. Keep the whole thing under 500 words so it fits on one page.
