Most business ideas don’t fail because the product’s bad, they fail because nobody cared enough to buy it at the price required. Market research is how you find that out early, before you burn cash, time and confidence.
If you’re still choosing what to build, cross-reference Business Ideas: The Full Guide to Finding, Testing and Choosing the Right Idea first, then use the process below to pressure test demand, competitors, pricing and buying behaviour.
In this article, we’re going to discuss how to:
- Pin down the exact customer problem and the job you’re solving
- Pull demand, competitor and pricing signals you can trust
- Run small tests in 7 to 14 days to prove buying behaviour before you build
Market Research, In Founder Terms
Market research is the process of collecting evidence that a specific group of people will pay you for a specific outcome, in a specific way, at a price that leaves you enough margin to operate.
It’s not a slide deck. It’s a set of artefacts you can point at when you’re about to commit:
- Demand proof: People are already searching, buying, complaining, switching or spending time trying to solve it.
- Competitive reality: You know who you’ll beat, and how you’ll explain that in one sentence.
- Price reality: You have a range, a floor, and a reason the numbers work.
- Buying behaviour: You know how decisions are made, who signs off, and what triggers action.
Completion check: if you can’t write down who buys, why they buy, what they pay, and how you’ll reach them, you haven’t done market research, you’ve done browsing.
How To Do Market Research For A Business Idea: Start With A Sharp Problem Statement
If you start with ‘I’m building an app for X’, you’ll end up collecting random facts that don’t help you decide. Start with the pain and the person, then work forwards.
Use this framing and don’t move on until it’s tight:
Problem statement: ‘[Customer] struggles with [pain] because [constraint]. They currently use [workaround], which costs them [money/time/risk].’
Then write your offer in one sentence. It forces clarity and it becomes the spine of your research.
One-sentence offer template: ‘We help [specific customer] get [measurable outcome] in [timeframe] without [major objection], using [simple mechanism].’
Example: ‘We help independent dental practices reduce no-shows by 20% in 30 days without discounting appointments, using automated confirmation and last-minute fill.’ Now your market research has a target.
Pull Fast Signals In A Few Hours (Internal First, Then Public)
Good operators don’t start on Google. They start inside their own world, then broaden out.
Internal Signals You Can Gather Today
If you’ve ever sold anything, worked in an industry, or run ads, you’ve got internal data. It’s messy, but it’s real.
- Past enquiries: Old DMs, emails, inbound calls, proposal requests. What did people ask for using their own words?
- Objection log: List the top 10 reasons people didn’t buy. If it’s all ‘timing’ and ‘budget’, your positioning is probably vague.
- Customer list: Who paid fastest, who churned, who was high maintenance? Market research isn’t only about demand, it’s about demand you can serve profitably.
Completion check: you should have at least 20 real phrases customers used. Those become your interview questions, ad copy and search terms.
Public Signals That Take 60 To 90 Minutes
Now validate that the problem exists beyond your personal bubble.
Use Google Trends for search demand patterns to sense check whether interest is stable, seasonal, rising, or collapsing. You’re not looking for perfection, you’re looking for obvious red flags like a 5-year downtrend or demand that only exists for 2 weeks a year.
Use industry and population context when it matters. If you’re targeting UK consumers by region, scan relevant local data such as UK Office for National Statistics datasets to sanity check how many potential buyers live where you plan to launch.
If you’re going B2B in the UK, check the density and reality of your target customers using Companies House company information. It’s a fast way to avoid building a ‘solution’ for an audience that’s too small or too fragmented.
Demand Research That Goes Beyond ‘Likes’
Demand isn’t what people say they’d buy, it’s what they already do to solve a problem. Your job is to spot behaviour that costs them money, time, reputation, or risk.
Here are practical demand signals you can collect this week:
- Switching behaviour: People leaving a competitor, cancelling a contract, or asking for alternatives.
- Workarounds: Spreadsheets, WhatsApp groups, DIY templates, virtual assistants, ‘we just do it manually’.
- Budget already assigned: They’re already paying for something adjacent, even if it’s the wrong tool.
- Urgency triggers: Deadlines, compliance, staff turnover, penalties, seasonality.
A simple rule: if the problem costs less than £50 a month to ignore, you’ll struggle unless you’re high volume, viral, or both.
Competitor Research: Find Your Real Alternatives
Founders often think competitors are companies that do the same thing. In reality, your competitor is whatever the customer would do instead of buying from you.
Build a competitor map with three buckets:
1) Direct competitors: Same audience, same job, similar delivery.
2) Indirect competitors: Different delivery, same outcome. For example: a freelancer instead of software, or an agency instead of a tool.
3) ‘Do nothing’: Manual process, internal admin, ignoring the issue until it bites.
Then answer four questions for the top 5 alternatives:
- Promise: What outcome do they claim, in plain language?
- Proof: What evidence do they show: case studies, reviews, numbers, logos?
- Price: What do they charge, and what’s actually included?
- Path to buy: Self-serve checkout, demo call, quote, procurement?
Completion check: you should be able to say ‘We’re different because…’ without using fluffy words like ‘quality’ or ‘better service’. If you can’t, your market research is telling you something uncomfortable: you don’t have an edge yet.
Pricing Research And Unit Economics That Hold At Small Scale
Pricing is market research. If your numbers only work once you hit 10,000 users, you don’t have a business idea, you have a hope.
Start with competitor pricing as a range, then pressure test willingness to pay with simple maths.
A Quick Unit Economics Pass You Can Do On A Notepad
Pick one proposed product and fill this in:
- Price: £P per month or £P per job
- Direct costs: payment fees, tools, delivery costs, fulfilment, contractor time
- Gross margin: (Price minus direct costs) divided by price
- Support load: hours per customer per month
Operator rule: if gross margin is below 60% for a service-light product, you’ll be on a treadmill. For a hands-on service, lower margins can work, but only if you can control time and scope.
Example calc: You charge £299/month. Tools and transaction fees are £29. You spend 1.5 hours/month supporting a client at an internal cost of £40/hour, that’s £60. Direct cost total £89, gross margin £210, margin is 70%. Now ask the real question: can you acquire that customer for less than a few months of contribution margin?
Pricing Questions That Produce Real Answers
Don’t ask ‘what would you pay?’. Ask these instead during interviews or sales calls:
- Anchor: ‘What are you paying today, all-in, to handle this?’
- Consequence: ‘What happens if you don’t fix it in the next 30 days?’
- Trade-off: ‘If we could remove X, what would you stop paying for?’
This is where you learn whether the business idea has room for profit, not just interest.
Research Buying Behaviour, Not Opinions
When people say ‘that’s interesting’, they mean nothing. Buying behaviour is about process, constraints and triggers.
In your market research, capture:
Decision chain: User, influencer, buyer, approver. In small firms it might be one person, in larger ones it’s never one person.
Buying moment: What event forces a decision? A chargeback spike, a landlord inspection, a staff resignation, an audit, a new baby, a new contract.
Sales friction: Do they need a demo, a trial, procurement, a contract review, or a supplier onboarding form?
Time-to-value: How quickly do they need a win? If they need results in 48 hours, your onboarding needs to be sharp.
Completion check: you should be able to diagram the buying process in 6 boxes or fewer. If it takes 15 boxes, you’re either targeting the wrong segment or selling the wrong way.
A 7 To 14 Day Validation Path That Uses Market Research Properly
Market research should lead to action. Here’s a simple path that doesn’t require you to build the full product.
Days 1 To 2: Build The Evidence Pack
Create a one-page ‘evidence pack’ in Google Docs:
- The offer: Your one-sentence offer and who it’s for
- The proof: 3 competitor screenshots, 5 customer quotes, 10 search phrases
- The pricing hypothesis: 2 to 3 price points and what’s included
Days 3 To 6: Run 10 Customer Conversations
Do 10 short calls with your exact buyer profile. You’re aiming for patterns, not permission.
Rules that keep it honest:
- Record and transcribe: Write down exact phrases, they’re gold
- Talk about the last time: ‘Tell me about the last time this happened’ beats hypotheticals
- Ask for a next step: A pilot, a paid setup fee, an intro to the decision maker
Completion check: you want at least 3 people to request a follow-up or ask how it works. If nobody does, the pain might be real but your positioning is off, or the segment is wrong.
Days 7 To 14: Run A Small Paid Test
Pick one channel and spend just enough to learn. You’re validating demand and message, not scaling.
Three practical options:
- Landing page plus pre-order: Take payment for a beta, even if delivery is manual at first
- Paid discovery offer: Charge £100 to £500 for an audit, plan, or setup, then roll that fee into the first month
- Outbound with a calendar link: 50 targeted messages, track reply rate and booked calls
Operator thresholds that are useful early:
- Cold outbound: 8% to 15% reply rate is a sign you’ve got a real hook
- Landing page: 2% to 5% conversion to email signup from targeted traffic is a decent starting point
- Paid beta: 3 to 5 paid customers beats 300 ‘interested’ signups
This is how to do market research for a business idea that forces reality. You either get money, or you get a clear reason why not.
Operational Guardrails That Protect Margin And Your Calendar
Market research often ignores operations, then founders wonder why the business is stressful. Build guardrails into the offer before you scale anything.
Practical guardrails that work across most models:
- Define the ‘done’: What is included, what is out of scope, and what counts as success
- Standardise onboarding: A checklist, a single intake form, a 15-minute kickoff
- Limit support: One channel, stated response times, no endless WhatsApp firefighting
- Protect your best customers: Don’t let the noisiest 10% eat 50% of your time
A simple check: if you can’t deliver the first 10 customers with one person working 40 hours a week, the model will punish you as soon as it gets traction.
Three Micro Cases To Show What ‘Good’ Looks Like
Micro case 1: Local trades lead-gen with a tighter segment
A founder wanted ‘marketing for plumbers’. Research showed emergency callouts were the real profit centre. He positioned around ‘same-day emergency jobs’, ran 30 outbound messages to firms advertising 24/7 service, booked 6 calls, closed 2 at £750 setup plus £350/month. Market research shifted the segment, not the channel.
Micro case 2: B2B compliance tool with a manual beta
A small team planned software for HR compliance reminders. Interviews revealed buyers wanted ‘audit readiness’ not reminders. They sold a £250/month manual service using templates and fortnightly check-ins, kept gross margin at 65%, then built software only once churn was under 5% monthly.
Micro case 3: Consumer subscription that died on price, not interest
A wellness box tested well in surveys. Paid tests showed 4% landing page conversion, but only 0.6% completed checkout at £39/month. Dropping to £25 fixed checkout but killed margin after shipping. The market research saved six months of fulfilment pain.
Common Risks And Simple Hedges
Most bad market research fails in predictable ways. Here are the ones I see most, plus how to hedge them.
Risk: You only speak to friendly people. Hedge: recruit interviewees from cold outreach, not your mates and followers.
Risk: You confuse a loud niche for a big market. Hedge: estimate reachable buyers in your first 6 months, then multiply by a realistic close rate. If the maths is ugly, accept it early.
Risk: You copy competitor features. Hedge: copy the customer’s desired outcome, then simplify delivery. Feature parity is a trap.
Risk: You price based on cost, not value. Hedge: tie price to what changes for the buyer: revenue gained, time saved, risk reduced.
Do And Don’t: The Operator Checklist
- Do write your one-sentence offer before doing any research
- Do collect 20 to 50 exact customer phrases and use them as your keywords and hooks
- Do treat pricing as a research question and test it with paid behaviour
- Don’t rely on surveys as your primary evidence
- Don’t ignore ‘do nothing’ as your biggest competitor
- Don’t build the full product before you’ve sold a manual version to a few customers
If you keep it this practical, you’ll quickly learn how to do market research for a business idea in a way that changes decisions, not just documents them.
Download The Customer Interview Script Pack And Get Better Answers Fast
If you want to speed up the most valuable part of market research, the conversations, download the Customer Interview Script Pack: Ask the Right Questions Before You Build and use it to run 10 buyer calls this week. You’ll get cleaner insights on triggers, budgets and objections, and you’ll stop collecting ‘nice idea’ feedback that never turns into revenue.
- Start with a tight problem statement and a one-sentence offer, then collect evidence that matches it.
- Validate demand with behaviours and small paid tests, and make sure your unit economics work before you scale.
- Build operational guardrails early so the first 10 customers don’t destroy your margin or your calendar.
FAQs For Market Research For A Business Idea
What’s the difference between market research and validation?
Market research is gathering evidence about demand, competitors, pricing and buying behaviour. Validation is using that evidence to run tests that prove people will pay, ideally within 7 to 14 days.
How many customer interviews do I actually need?
Do 10 to start, then look for repeated patterns in language, triggers and objections. If your interviews are all over the place, your segment is too broad or your offer is unclear.
What if there are loads of competitors?
Competition usually means money is being spent, which is a good sign. Your job is to pick a narrower segment or a sharper outcome where you can win without feature parity.
Can I do market research without paying for tools?
Yes, you can get far with customer calls, competitor pricing pages, public datasets and a basic landing page. Spend money only when it helps you test behaviour, like a small ad budget or a paid outreach tool.
How do I research pricing if customers won’t tell me what they’ll pay?
Ask what they pay today and what it costs them to ignore the problem, then test pricing with a paid beta or paid discovery offer. Real pricing data comes from transactions, not opinions.
What’s a good early sign my idea has demand?
People describe the problem quickly, mention a recent incident, and ask what the next step is. The strongest signal is someone paying for a manual version even before the product exists.
How do I avoid building something that’s unprofitable?
Run a simple unit economics pass early, including your time cost, and set a gross margin target you won’t break. If the maths only works at scale, change the offer, the segment, or the delivery model.
When should I stop researching and start building?
Start building when you’ve sold to a few customers, or you’ve got a clear paid path to do so. If you’re still getting ‘interesting’ but no commitments, you need a tighter segment or a stronger offer before you write code.
