# The AI said my $58,000 grooming van pays off in 26 weeks. Local prices said 193. > An AI market scan told a groomer her new van would pay for itself in about half a year. We tested a How I AI workflow that makes the AI argue the other side first. By Jordan Menlo · Published October 24, 2026 Section: Markets & Main Street › Small Business · Series: AI Prompts, Reviewed (https://mcafax.com/notebook/prompts) Canonical: https://mcafax.com/notebook/ai-prompt-review-market-research Source: MCAFax, The Broker's Notebook (https://mcafax.com/notebook). MCAFax is not a lender or a law firm. --- "Strong opportunity. Demand for mobile grooming is growing fast in your area." That's what **Tess, owner of Quillfeather Grooming (a made-up dog groomer)**, got back when she asked an AI whether to buy a used grooming van. It even did the math for her. The van would pay for itself in about 26 weeks. She almost signed. Then she checked the AI's sources. If the AI finds a market for every idea, make it argue the other side first. **AI Prompts, Reviewed** takes one prompt or workflow from How I AI, Claire Vo's show where guests share their screens and walk through how they really use AI. We run it on a small-business job, score it, and tell you what we'd change. **Credit:** Based on "How to Conduct Rapid Market Research and Competitive Analysis Using AI," a workflow from How I AI (hosted by Claire Vo, written up by the ChatPRD team), shown by Anjan Panneer Selvam, CPTO of Acolyte Health, in the episode "How I AI: Anjan Panneer Selvam and a new model of B2B product management." [Read the original workflow on How I AI](https://www.chatprd.ai/how-i-ai/workflows/how-to-conduct-rapid-market-research-and-competitive-analysis-using-ai). [Watch the full episode](https://www.chatprd.ai/how-i-ai/b2b-product-management-prototypes). Guest site: none listed on the How I AI workflow page. ## What the original does In the episode, Anjan runs a product idea through Perplexity, turns the result into a Gamma slide deck, then asks the AI to play devil's advocate. He says that step has helped him say no to more features than yes. The written workflow has 4 steps: - Write the idea as a research brief: customer, job, current alternative, location, business model, and the decision it should inform. - Run a cited scan that compares competitors, substitutes and workarounds on the same points, and splits fact from guess. - Build a slide outline with one claim per slide and its source, then let the deck tool lay it out without adding numbers. - Show it to the people deciding, and write down what would make you stop, narrow or go ahead. The line we'd tape to the monitor: **"End with the strongest case against pursuing it."** ## The test Everything here is made up. Quillfeather is a one-location groomer, and Tess wants a van for clients who can't drive in. - Van: a used, fitted grooming van for **$58,000** - Running costs: about **$1,400** a week for the groomer's pay, fuel, insurance and supplies **Run 1, the shortcut:** "Is mobile dog grooming a good opportunity in my town? Should I buy a van?" The answer was all upside. It listed **8** mobile competitors as proof of demand, said a mobile groom goes for **$120**, and planned on **6** grooms a day, **5** days a week. Its math: **6 x $120 = $720** a day, or **$3,600** a week. Take off the **$1,400** in running costs and that's **$2,200** a week. The van divided by that, **$58,000 / $2,200**, is about **26** weeks. ## The rerun **Run 2:** same tool, our prompt below. It forces sources with dates, the same comparison points for every competitor, and a case against. Tess opened every link. Of the **8** "competitors," **3** had closed in the past year and **2** were regular shops, not vans. That left **3** real mobile groomers nearby, and the closed ones were a warning, not proof of demand. The **$120** price came from a big-city groomer's page. The **3** local vans listed a basic groom at **$85**. The case against also flagged drive time. With travel between homes, **4** grooms a day was the realistic number, not **6**. New math: **4 x $85 = $340** a day, or **$1,700** a week. After the same **$1,400** in costs, that's **$300** a week. Now **$58,000 / $300** is about **193** weeks, which is more than three and a half years. Tess didn't buy the van. Her next step is a cheap test: rent a fitted van one day a week for a month and see if her own clients book it. ## Verdict: 8/10 **Strengths:** it makes the AI separate fact, guess and unknown. Comparing every competitor on the same points catches mixed-up lists. And the case-against step is the one most owners skip. **Weakness:** it's built for a product team with a deck and a room of stakeholders. Owners need one honest page and a payback number they can trust, and the workflow leaves that math to you. **Founder tweak:** feed it your own local prices and costs, make it show the payback math line by line, and click every source before you believe a number. ## Copy the MCAFax version This is our own rewrite, not the original. [**Copy the original from Anjan Panneer Selvam on How I AI**](https://www.chatprd.ai/how-i-ai/workflows/how-to-conduct-rapid-market-research-and-competitive-analysis-using-ai)**.** ``` I own [business type] in [town and state]. I'm deciding whether to [decision, e.g. buy a $X van]. Use only information you can link to, and give the date for each source. Step 1. List competitors within [distance] of [town]. Put each one in a group: same service, different service (substitute), or a workaround customers use instead. For each, show: what they offer, their posted price, where you found it, and the source date. Mark any business that looks closed or moved. Step 2. Mark every claim as FACT (linked), GUESS (your estimate), or UNKNOWN. Step 3. Do the money math line by line with these inputs: Cost: [$]. Weekly running costs: [$]. My price: [$ or local average]. Jobs per day: [number, counting travel or setup time]. Days per week: [number]. Show revenue, profit per week, and weeks to pay back the cost. Step 4. Give me the strongest case AGAINST doing this. Include businesses like mine that closed, what would make customers not switch, and anything that could cut the jobs per day. Step 5. End with one cheap test I can run in 30 days before I spend the money. Do not invent prices, statistics, or competitors. If you can't find it, say UNKNOWN. ``` Paste into an AI tool that can search the web and show its sources. ## Before you spend the money 1. Click every source, and check that the business is open and nearby. 2. Use local posted prices, not a national or big-city number. 3. Count travel or setup time when you set jobs per day. 4. Run the payback math yourself with a calculator. 5. Keep watching what nearby competitors charge with our [competitor price monitoring prompt](https://mcafax.com/notebook/prompts/competitor-price-monitoring). More owner prompts: [browse the MCAFax prompt library](https://mcafax.com/notebook/prompts). ## Questions readers ask **What tools does the original workflow use?** Perplexity for the cited research scan and Gamma to turn a source outline into a slide deck. **Why did the first answer look so good?** A loose question invites a sales pitch. AI tends to find a market for every idea unless you ask it for the case against. **What is a devil's advocate pass?** A follow-up that asks the AI for the best reasons not to go ahead, such as closed competitors, switching costs or weak demand. **Do I need to make a slide deck?** No. For a small business, one page with the sources, the payback math and the case against is enough. **How do I check the AI's competitor list?** Open each link, confirm the business is open and nearby, and make sure it sells the same service at a posted price. **Can I trust the payback number?** Only after you swap in your own prices and costs and redo the math. The AI's inputs are guesses until you check them.