A complete afternoon workflow that gives you a competitor audit, audience analysis, and gap analysis using free AI tools. Specific prompts, specific outputs, zero cost.
Market research used to mean one of two things. Hire a research firm for £10,000. Or spend three months designing surveys, running focus groups, and assembling findings into a report that's outdated by the time you read it.
Neither option works when you're a solopreneur or small business trying to figure out whether your offer actually fits the market. You need answers this week, not this quarter. And your budget is closer to £0 than £10,000.
Here's what's changed. You can now run a genuinely useful competitor audit, audience analysis, and gap analysis in a single afternoon using free AI tools. Not surface-level guesswork. A structured AI market research workflow that produces a positioning document you can act on immediately.
This is the exact process. Four connected steps, about two and a half hours total, and a finished one-page market positioning document at the end.
Forget everything you've read about market research methodology. Focus groups, sample sizes, confidence intervals — that's for Unilever launching a new shampoo into 40 countries. You're trying to figure out whether your offer is positioned correctly and where the gaps are.
For a small business, market research answers three questions:
That's it. Three questions. If you can answer them with evidence — not gut feeling — you have enough market research to make real decisions about your positioning, your messaging, and your offer.
The workflow below answers all three in one session. Each step feeds the next, so by the time you finish, the research has already been synthesised into something actionable.
This is a connected four-step process, not a list of isolated tasks. The output from each step becomes the input for the next. Here's the structure:
Total: about two and a half hours. The output: a one-page market positioning document with evidence behind every claim.
Let's walk through each step with specific tools, specific prompts, and specific outputs.
Pick five competitors. Not five aspirational brands you admire. Five businesses that your target customers would actually consider instead of you. If you're not sure who they are, Google the service you offer in your area or niche. The businesses on page one are your competitors.
For each competitor, you need to understand four things:
Visit each competitor's website. Copy their homepage text, about page, and pricing page into a document. Then feed it to ChatGPT with this prompt:
"I'm analysing five competitors in [your niche]. Here's the homepage and pricing content from each one. For each competitor, identify: (1) their core positioning statement, (2) the primary audience they're targeting, (3) their pricing model and price points, (4) three strengths in their messaging, (5) three weaknesses or gaps in their messaging. Then create a comparison table showing where they overlap and where they differ."
This single prompt gives you a structured competitive landscape in minutes. But don't stop there.
For each competitor, open Google and search site:competitor.com. This shows you every page Google has indexed. You'll see their blog posts, landing pages, and resource pages — their entire content strategy laid bare.
Scan the titles. What topics come up repeatedly? What are they trying to rank for? Where are they investing their content effort? Copy the top 10–15 page titles from each competitor into your document.
Then feed those to ChatGPT:
"Here are the blog post titles from five competitors in [your niche]. Identify: (1) the topics every competitor covers, (2) topics only one or two competitors cover, (3) obvious topic gaps that none of them are addressing."
By the end of this step, you have a clear picture of what your competitors are saying, how they're saying it, and where they're not saying anything at all. If you want to go deeper on this step, the AI competitor analysis for small business guide covers the full process.
A competitor comparison table showing positioning, pricing, messaging strengths, messaging weaknesses, and content gaps for five competitors. Save this — you'll need it for Step 3.
Competitor analysis tells you what businesses are saying. Audience research tells you what customers are actually thinking. These are almost never the same thing.
The best audience research doesn't come from surveys. It comes from places where people complain in their own words, unprompted, to strangers on the internet.
You're looking for two things: pain points and language.
Pain points are the specific frustrations. Not "they want better marketing." Something like "I've tried three different social media schedulers and none of them actually help me know what to post." That's a pain point you can build an offer around.
Language is how they describe those frustrations. The exact words and phrases they use. This is more valuable than any copywriting course because it's your audience telling you how to talk to them.
Collect 10–15 real quotes from these sources. Direct quotes, copied verbatim. Then feed them to ChatGPT:
"Here are 15 real quotes from [your audience] taken from Reddit, Facebook groups, and Amazon reviews about [your topic]. Analyse these quotes and identify: (1) the three most common pain points, (2) the three most common desired outcomes, (3) recurring phrases and language patterns, (4) any frustrations with existing solutions. Use only the evidence in these quotes — don't infer anything that isn't directly stated."
This is more valuable than any survey. Surveys tell you what people think they should say. Forum posts tell you what they actually feel. The constraint in that prompt — "use only the evidence in these quotes" — stops ChatGPT from hallucinating patterns that aren't there.
If you want to turn this audience data into a detailed customer profile, the buyer persona with AI guide shows you how.
A document listing the top pain points, desired outcomes, recurring language, and frustrations with existing solutions — all backed by real quotes. Save this for Step 3.
This is where the workflow pays off. You now have two documents: what competitors are offering (Step 1) and what customers actually want (Step 2). The gap analysis overlays one on top of the other.
The gaps — the places where customer needs aren't being met by existing solutions — are your opportunity.
Take your competitor comparison table from Step 1 and your audience analysis from Step 2. Feed both into ChatGPT:
"I have two documents. Document 1 is a competitor analysis of five businesses in [your niche], showing their positioning, pricing, messaging, and content strategy. Document 2 is an audience analysis based on real customer quotes, showing pain points, desired outcomes, and frustrations with existing solutions. Compare these two documents and identify: (1) customer needs that no competitor is addressing well, (2) pain points that competitors acknowledge but don't solve, (3) language and framing that customers use but competitors don't, (4) pricing or packaging gaps — things customers want to buy that nobody is selling in the right format. Rank the gaps by potential impact."
This prompt is doing something a human researcher would take days to do manually. It's cross-referencing supply (what's available) against demand (what's needed) and identifying the mismatches.
Not every gap is a real opportunity. Some gaps exist because the market already tried to fill them and it didn't work. Some gaps are too small to matter. So challenge each one.
For each gap ChatGPT identifies, ask:
You can ask ChatGPT to play devil's advocate: "For each gap you identified, give me three reasons it might not be a viable opportunity." This forces the analysis to be honest rather than optimistic.
A ranked list of 3–5 market gaps with evidence for each one — which customer pain points it addresses, which competitor weaknesses it exploits, and preliminary viability notes. This feeds directly into Step 4.
You have a hypothesis. "There's a gap in the market for X." Now you need to check whether real demand exists. Gut feelings aren't market research. Neither are AI-generated analyses. This step grounds everything in actual data.
Open Google Trends and search for terms related to your identified gaps. You're looking for two things: Is interest growing, stable, or declining? And is there enough interest to sustain a business?
A topic with steady or rising interest over the past 12 months is a green light. A topic that peaked two years ago and is in freefall is a red flag — no matter how good the gap analysis looks.
Compare your gap terms against your competitors' positioning terms. If your gap has comparable or growing interest relative to what competitors are already targeting, that's a strong signal.
Open Keyword Planner (free with a Google Ads account — you don't need to run ads). Search for phrases related to your gaps. You want to see monthly search volume and competition level.
Look for keywords with 100–1,000 monthly searches and low to medium competition. That's the sweet spot for a small business. High-volume, high-competition keywords are battlegrounds for big brands. Low-volume, low-competition keywords might not have enough demand. The middle ground is where solopreneurs win. The AI keyword research workflow covers this in full detail.
Post a quick poll or question on your social media or in a relevant community. Don't pitch your solution. Ask about the problem.
"What's the most frustrating part of [topic] for you?" or "If you could get help with one thing related to [topic], what would it be?"
You don't need 500 responses. Even 10–15 genuine answers will either confirm your gap analysis or reveal that you've missed something. The point is to test your research against real people before you build anything.
A validated shortlist of 1–3 market gaps with evidence from Google Trends (demand trajectory), Keyword Planner (search volume), and social responses (qualitative confirmation). You now have a market positioning hypothesis backed by data at every layer.
The AI Content System connects market research to content strategy to published posts — one connected workflow.
Get the AI Content System — $29Generic prompts produce generic output. "Analyse my market" will give you a Wikipedia-grade summary that helps nobody. The prompts that work share three qualities: context, constraints, and format requirements.
"I run a [your business type] targeting [your audience]. Here are the homepage texts from five competitors: [paste texts]. For each competitor, identify their core value proposition in one sentence, the primary objection they're trying to overcome, and the emotional trigger they're using. Present this as a comparison table with competitors as rows and these three elements as columns."
Why it works: It gives ChatGPT a specific framework (value prop, objection, emotional trigger), asks for a specific format (comparison table), and provides real data to analyse rather than asking it to guess.
"Here are 12 real customer quotes from Reddit and Amazon reviews about [your topic]. Extract: (1) the exact phrases they use to describe their problem — not your interpretation, their words, (2) the solutions they've already tried and why those didn't work, (3) what their ideal solution would look like based on what they're saying. Do not add any information that isn't directly supported by these quotes."
Why it works: The constraint — "do not add any information that isn't directly supported by these quotes" — is critical. Without it, ChatGPT will embellish, generalise, and add plausible-sounding patterns that don't actually exist in the data. With it, you get honest extraction.
"Based on these competitor offerings [paste summary] and these customer pain points [paste summary], identify the top five underserved needs. For each one, explain: (1) which customer pain points it addresses, (2) why current competitors aren't solving it, (3) what a solution would need to include to fill this gap effectively. Rank them by the size of the gap — not by how easy they are to fill."
Why it works: It explicitly asks for ranking criteria (size of gap, not ease) and requires evidence linkage (which pain points, which competitor weaknesses). This stops the output from being a vague wish list.
"I've identified these three market gaps: [list them]. For each one, give me three specific reasons it might not be a viable business opportunity. Consider: market size limitations, execution difficulty, reasons competitors may have intentionally avoided this space, and whether the pain point is severe enough for people to pay to solve it."
Why it works: Most people only ask AI to confirm their ideas. This prompt forces it to challenge them. If the gaps survive this scrutiny, they're worth pursuing. If they don't, you've saved yourself months of building the wrong thing.
The entire workflow runs on free tools. Here's the complete toolkit:
site:competitor.com into Google to see every page they've published. Free.That's five tools. All free. No subscriptions. No trials that expire in 14 days. This is the whole stack you need for AI market research that produces actionable results.
If you're building out a full marketing toolkit alongside this, the step-by-step AI marketing system shows how market research connects to content, SEO, and lead generation.
Research is worthless if it stays in a ChatGPT conversation. The single most important thing you can do after finishing this workflow is extract the findings into a document you'll actually use.
One page. Not a 40-slide deck. One page with these sections:
Print it. Pin it to your wall. Reference it every time you write a blog post, create an ad, or update your website. Every piece of marketing you produce should be informed by this document.
The audience research in Step 2 gave you real customer language. Use it. If your customers describe their problem as "I don't know what to post," your headline shouldn't say "comprehensive social media content strategy solutions." It should say "know exactly what to post." Match their language.
The competitor content audit in Step 1 showed you what topics are being covered and what's missing. The gap analysis in Step 3 showed you where the underserved needs are. Your content calendar should target those gaps directly. Write about the things nobody else is writing about, for the audience nobody else is serving properly.
If the research reveals that your audience wants something different from what you're selling, adjust. Maybe you're selling a course when they want a template. Maybe you're pricing monthly when they want a one-off. The research tells you. Listen to it.
This isn't theoretical. Here's how we used this exact workflow to research and position Syxo.
We analysed the AI marketing education space. Dozens of courses, YouTube channels, newsletters, and communities. The competitor audit revealed a pattern: almost everyone was selling prompt collections. "500 ChatGPT prompts for marketers." "The ultimate prompt library." Prompts were the product.
Positioning was remarkably similar across the board. "Use AI to save time on marketing." Pricing clustered around either free (ad-supported content) or premium ($200+ courses). The mid-range was empty.
We read hundreds of posts in Reddit threads, Facebook groups for small business owners, and indie hacker communities. The pattern was unmistakable.
People weren't complaining about a lack of prompts. They had plenty of prompts. They were complaining that the output was generic. "I used ChatGPT and it gave me something that sounds like everyone else." "The prompts work but the results don't feel like my brand." "I can get AI to write a post but I don't know what to write about in the first place."
The pain wasn't prompts. It was systems. People didn't know how to connect the steps — from research to strategy to content to publishing. They had individual tools but no workflow.
When we overlaid the competitor offerings (prompts, isolated tools, generic advice) against the customer complaints (output sounds generic, no connected workflow, no system), the gap was clear.
Nobody was selling connected marketing systems — workflows where the output of one step feeds the input of the next, calibrated to a specific business and voice. Everyone was selling components. Nobody was selling the architecture.
Google Trends showed steady growth in "AI marketing system" and "AI marketing workflow." Keyword Planner confirmed search volume across dozens of system-related keywords. Community polls confirmed the frustration: people wanted to know how the pieces fit together, not just what the pieces were.
That research became Syxo's positioning: "Systems, not prompts." Every product, every blog post, every piece of content maps back to that one insight from an afternoon of research.
Yes. ChatGPT's free tier, Google Trends, Google Keyword Planner (free with a Google Ads account), and public forums like Reddit and Quora give you enough data for a solid competitor audit, audience analysis, and gap analysis. Paid tools add speed and depth, but the free stack covers the essentials.
With a structured workflow, about two and a half hours. That breaks down to 45 minutes on competitor analysis, 30 minutes on audience research, 30 minutes on gap analysis, and 30 minutes on validation. The output is a one-page positioning document you can act on immediately.
AI market research is a strong starting point, not a final answer. The competitor audit and audience research steps use real public data — actual customer complaints, actual competitor positioning, actual search volumes. The AI helps you synthesise patterns faster. Always validate findings with real market signals before making major decisions.
ChatGPT is the most versatile free option for market research because it can analyse competitor positioning, identify patterns in customer feedback, and synthesise gap analyses. But the real power comes from combining it with Google Trends for demand validation and Google Keyword Planner for search volume data. No single tool does everything.
The AI Content System connects market research to content strategy to published posts — one connected workflow.
Get the AI Content System — $29