How to do market research, step by step
Market research comes down to five steps: define the market and put a number on it, find the data, analyze the demand, analyze the supply, conclude. Plan for 20 to 40 hours if you do it yourself, or $2,500 to $40,000 if you hand it to someone in Quebec. The free sources that matter here are Statistics Canada, the Institut de la statistique du Québec, the Registraire des entreprises and the paid databases you can open for free with a BAnQ library card.
- The five steps are a sequence, not a list: putting a number on the market before surveying keeps you from surveying for nothing.
- A BAnQ library card gives free access to industry databases that cost thousands of dollars per subscription.
- Twelve interviews are enough for most SMB markets: beyond that, you hear the same answers.
- A study that cannot conclude “no” is not a study, it is a justification.
- Monthly search volumes are the fastest and cheapest source for measuring demand, and almost nobody uses them.
Updated August 27, 2026
On this page
What market research is, and what it is not
Market research is verification work: it establishes whether enough people want what you intend to sell, at a price that leaves you a margin, in a territory you can reach. It produces a potential revenue figure and a decision, not a document.
It is not a writing exercise to satisfy a financial institution, even if that is often what brings you here. It is not a survey either: the survey is one tool out of five. And it is not a feasibility study, which deals with your ability to produce, whereas market research deals with other people's willingness to buy.
Market research answers a single question: how many people, ready to pay how much, can be reached at what cost. Anything that does not serve that question makes the document heavier without making it better.
How long market research takes and what it costs in Quebec
From 20 to 40 hours if you run it yourself, $2,500 to $8,000 with an independent consultant, $10,000 to $40,000 with a research firm. The gap does not come from the quality of the analysis, it comes from primary data collection: recruiting and questioning respondents is expensive, and it is the only line item that balloons.
| Approach | Your time | Your money | What you get | The right choice when |
|---|---|---|---|---|
| Yourself | 20 to 40 hours | $0 to $500 | Public data, 10 to 15 interviews conducted by you, competitor analysis | Local market, tight budget, you already know the sector |
| Independent consultant | 5 to 10 hours | $2,500 to $8,000 | The same thing, done faster and by someone outside the project | You are short on time or you need a view that is not your own |
| Research firm | 3 to 6 hours | $10,000 to $40,000 | Quantitative survey with a panel, recruited focus groups, formal report | Major investment, new market, or a funder's requirement |
The timeline roughly follows the budget. Plan for three to six weeks doing it yourself at a few hours a week, two to four weeks with a consultant and six to ten weeks with a firm, three of which go to recruiting respondents alone.
Doing it yourself
This is the right option for most Quebec SMBs, and not only for budget reasons. You know the sector, you have access to potential customers nobody will recruit better than you, and you will hear things in the interviews that a report never conveys. The risk is real: you are looking to confirm what you hope for. The safeguard is written further down, in the section on walk-away thresholds.
An independent consultant
Between $2,500 and $8,000 depending on scope, generally without quantitative data collection. You are mostly buying time and distance. Ask three things before signing: how many interviews are included, who conducts them, and whether the report can end with a recommendation to walk away.
A research firm
From $10,000 to $40,000. What you are paying for is data collection: a survey of a representative panel is expensive because every respondent is paid, and a recruited focus group costs $3,000 to $6,000 per group, room and moderation included. This level is justified when the decision commits several hundred thousand dollars, or when a funder explicitly requires it.
A good proposal says how many respondents, recruited how, and at what margin of error. A vague proposal talks about a “representative sample” without giving a number.
Step 1. Define the market and put a number on it
Start with the number, not the questionnaire. Sizing the market before surveying saves you from spending three weeks questioning people about a market that, once the math is done, cannot support your project. It is the step most people skip, and it is the one that makes the others fail.
Total market, available market, attainable share
Three nested circles, each one calculated from the one before. The total market is the annual spend of everyone who buys this type of product. The available market is the portion your offer and your territory can actually serve. The attainable share is what you can hope to take in three years, given the competition and your means.
The attainable share is almost always overestimated. A new entrant in an established market rarely takes more than 1 to 3% in three years, unless it arrives with a price or distribution advantage nobody else has. If your calculation gives 10%, redo it by writing down the names of the customers you are going to take from your competitors.
A complete calculation, with numbers
Take a concrete case: a commercial cleaning service that wants to start up in the Laurentians. Here is the calculation from end to end, with the source of each number. Replace the values with your own: the structure does not change.
| Line | Number | Where it comes from |
|---|---|---|
| Commercial establishments in the Laurentians | 4,800 | Statistics Canada, Business Register, by region and by size |
| Minus those with fewer than 5 employees, who do the cleaning themselves | 2,100 remaining | Same source, filtered by size |
| Average annual spend on cleaning | $6,200 | Three comparable quotes obtained from competitors |
| Total market | $13,020,000 | 2,100 × $6,200 |
| Territory actually served, a 30-minute drive | 38% of the total | Manual count on a map |
| Available market | $4,947,600 | $13,020,000 × 38% |
| Share targeted in three years | 2% | Cautious new-entrant assumption |
| Potential revenue, year 3 | $98,952 | $4,947,600 × 2% |
That figure, about $99,000 in annual revenue by the third year, is what the study has to produce. It changes everything: it says how many sales need to be closed, at what pace, and whether the project can support one person or three. Without it, the rest of the study has nothing to validate.
Step 2. Where to find the data in Quebec
Almost everything you need is public and free. This is the most serious gap in existing guides: they explain that you need secondary data without ever saying where to click. Here is the list, in the order you open them.
Free public sources
Statistics Canada
The Census Profile gives the population, age, income and education level of any Quebec municipality. The Business Register gives the number of establishments by sector, by region and by employee bracket: it is the source of the first number in any B2B market calculation. The tables on business dynamics give survival rates by sector, which is infinitely more useful than the bankruptcy statistic that circulates without a source.
The Institut de la statistique du Québec
The Quebec data that Statistics Canada does not have or publishes later: population projections by regional county municipality (MRC), disposable income by region, sector surveys on construction, tourism, culture and agri-food.
The Registraire des entreprises du Québec
Free and often overlooked. For each competitor, you will find its registration date, its legal form, its shareholders, its establishments and its declared number of employees. It is the fastest way to find out whether the competitor that worries you is a company of three people or thirty.
Données Québec and municipal open data
Building permits, assessment rolls, foot traffic, transportation data. Useful mainly for projects tied to a physical location.
Paid databases you can open for free with a BAnQ library card
A Bibliothèque et Archives nationales du Québec library card is free for every Quebec resident and can be obtained online. It gives remote access to commercial databases whose individual subscription costs thousands of dollars a year: industry reports, company profiles, financial ratios by sector, trade journals. It is the most underused tool in the whole process.
What you look for there, concretely: the average financial ratios for your sector, which tell you what margin a typical competitor earns, then the industry reports, which give you the size of the national market to break down by region afterward.
Industry associations
Almost every Quebec sector has an association that publishes an annual profile: number of companies, sector revenue, trends, salaries. These documents are often free or available for the price of an annual membership, and they contain figures you will find nowhere else. Search for the name of your sector followed by “association Québec portrait”.
What search volumes tell you
This is the fastest, cheapest and most neglected method. The number of monthly searches on a query measures real, expressed demand, without anyone having to answer a questionnaire. Google's Keyword Planner is free with an advertising account, and Google Trends is free without one.
| What you look at | What it tells you | The warning sign |
|---|---|---|
| Monthly volume on your commercial queries | How many people are actively looking to buy | Fewer than 50 searches a month in total across all your queries |
| Five-year trend in Google Trends | Whether demand is rising, flat or collapsing | A curve in steady decline for three years |
| Your competitors' Google reviews | What really frustrates their customers, in their own words | No competitor has any reviews: nobody looks for this product online |
| Your competitors' job postings | Whether they are growing, and in which function | No posting in two years across the whole sector |
| Local groups on social media | The questions asked before buying, for free | The subject never comes up |
Demand that does not exist in search volumes can still exist: some products sell through calls for tenders, referrals or outbound prospecting. But if you were counting on the web to sell, a volume of zero is an answer in itself.
Step 3. Analyze the demand
You are looking for three things: who buys, what they do today in the absence of your offer, and how much they pay for that substitute. The third is the most important and the least often asked.
How many people to interview before it counts
For qualitative interviews with a homogeneous customer base, saturation arrives at around twelve interviews: from that point on, you hear the same answers. It is a threshold documented in the methodological literature, and it holds for the majority of SMB projects. If your customer base splits into truly different groups, plan for twelve interviews per group, not twelve in total.
| Method | How many people | Margin of error | What it costs |
|---|---|---|---|
| Qualitative interviews | 12 per segment | Not applicable, it is not a measurement | Your time, about 45 minutes each |
| Focus group | 5 to 10 per group, 2 to 3 groups | Not applicable | $3,000 to $6,000 per group if recruited by a firm |
| Survey, rough precision | 100 respondents | About ± 10 points | Doable on your own with a free form |
| Survey, usual precision | 400 respondents | About ± 5 points | Paid panel from $4,000 |
| Survey, fine precision | 1,000 respondents | About ± 3 points | Paid panel from $10,000 |
On response rates, be realistic. An email sent to a cold list gets 2 to 5% responses. The same request sent to your own customers gets 10 to 25%. To get 400 respondents from a cold list, you therefore have to write to more than ten thousand people, which explains the price of a panel.
A 12-question interview script, ready to copy
The first eight minutes are about what the person does today, never about what they would do. Do not present your offer until question 9, otherwise everything before it is contaminated by politeness.
Tell me about the last time you had this problem
An open question, with no mention of your product. Let them talk.
What did you do at that point?
You are looking for the real substitute, which is often “nothing” or “we managed”.
How many times a year does this happen to you?
Frequency determines whether you are selling a subscription or a transaction.
How much did it cost you, in money and in time?
The figure the person gives here is the ceiling of your price.
Who made the decision, and who else had a say?
In B2B, this question saves months of prospecting aimed at the wrong person.
How did you find the supplier?
You are looking for the real channel: online search, referral, call for tenders, advertising.
What made you choose that one over another?
The real decision criterion, which is rarely price.
What disappointed you most?
This is where your position in the market is found.
If someone offered this, what would you ask yourself before buying?
First mention of your offer. Framed as objections, not as interest.
At what price would you find it too expensive? And at what price would you find it suspicious?
Two limits are better than one question about an acceptable price, to which everyone answers “cheap”.
What would stop you from changing suppliers?
The switching cost is what protects your competitors, and it will protect you later.
Who else should I interview?
The question that fills your interview calendar. Never skip it.
What people say versus what they buy
A stated intention is not a sale. People say yes out of politeness, overestimate their budget and underestimate their inertia. Three ways to test real demand before investing: ask for a refundable deposit, have a letter of intent signed, or sell the product before you have it. A person who takes out their card is a data point; a person who says “good idea” is not.
The fastest and cheapest test is a single page describing the offer, with a form, pushed by $300 to $500 of targeted advertising over two weeks. You get a real cost per inquiry, from real people, in your territory. That figure is worth more than thirty interviews, because it measures an action instead of an opinion. And along the way it gives you the acquisition cost to put in your projections.
Step 4. Analyze the supply
Listing your competitors takes two hours and is done almost entirely online. What takes time is understanding what they do badly, because that is where you will get in.
Listing the real competitors
Three categories, and the third is the one that gets forgotten. Direct competitors, who sell the same thing. Indirect competitors, who solve the same problem another way. And the most common substitute of all: doing nothing, or doing it yourself. That last one often takes more market share than all the others combined, and it appears on no list of competitors.
For each one, note five things: its year of registration with the Registraire, its declared number of employees, its price as posted or obtained by quote, its average rating on Google, and what its three most negative reviews complain about.
Reading their Google reviews as a data source
A competitor's reviews are a free focus group that it paid for on your behalf. Read the three-star reviews before the one-star ones: one-star reviews often describe an isolated incident, whereas three-star reviews describe a structural disappointment, the kind you can fix in your own offer.
Note the exact words people use. They are the ones they will type into Google, and the ones your site will need to pick up.
What their job postings tell you
A company hiring three sales representatives is preparing an expansion. A company hiring a logistics director is changing scale. A company that has posted no position in two years, in a sector where the others are hiring, is probably in difficulty or for sale. Job postings are a leading indicator, and a public one, and nobody looks at them.
Step 5. Analyze the environment and conclude
Four factors to check, and a single one is enough to kill a project: the applicable regulations, the barriers to entry, dependence on a single supplier or customer, and sensitivity to the economic cycle. Take one hour per factor, no more.
In Quebec, two checks come up almost every time: the permit or certification your sector requires, and the language and personal information protection obligations if you sell online. Both can be dealt with, but they carry a cost that has to be calculated before, not after.
The conclusion fits on one page and contains four numbers: the potential revenue at three years, the investment needed to reach it, the cost of acquiring a customer, and the time to break-even. If one of those four numbers is only a hunch, the study is not finished.
What your conclusion must contain
A study that does not end with these four numbers ends with an opinion, and an opinion cannot be financed. The questions to ask yourself before saying the work is done:
- The potential revenue at three years, and the calculation that leads to it, line by line?
- The investment required, including the months with no revenue?
- The cost of acquiring a customer, measured and not estimated?
- The time to break-even, in months?
- The threshold below which you walk away, written before you have the results?
The six mistakes that invalidate market research
Interviewing people close to you
They want what is best for you, so they lie to you. No interview with someone who knows you counts toward the twelve.
Asking questions about the future
“Would you buy this?” measures nothing. “What did you buy last time?” measures a behaviour.
Confusing the total market with the attainable market
The Canadian market for your sector does not concern you if you serve three regional county municipalities. It is the mistake that inflates business plans and makes lenders smile.
Forgetting the “do nothing” substitute
Your main competitor is often the customer's inaction, and it is free.
Copying out-of-date data
Check the date of every figure. A 2019 industry report describes a world that no longer exists.
Writing the study after making the decision
It is the most common one, and the only one that makes all the others irrelevant.
How to know the answer is no
Write your walk-away thresholds before collecting anything, and date the page. It is the only protection against the urge to reinterpret the results once they are in. A study that cannot conclude “no” is not a study, it is a justification.
Five thresholds to write in advance
A “no” obtained for 40 hours of work is the best return of the whole exercise. It costs a week, and it sometimes saves two years.
The template, filled in with a complete example
We have put the template online, on one page that follows exactly the five steps above. Each section carries the questions to fill in, the sources to open and the worked example of the commercial cleaning service in the Laurentians, so you can see what a completed box looks like.
The page prints or saves as a PDF from your browser, with the print command. There is nothing to download, nothing to hand over and no form to fill out.
Open the market research template, on one page, with the full worked example.
Opening a new market is at the heart of the Develop a new market goal, which starts from exactly this work.
If your study concludes yes and you want an outside opinion on the entry strategy before committing the budget, the 90-minute consultation is made for that, and you leave with a written summary.
Frequently asked questions
How much does market research cost in Quebec?
From $0 to $500 if you run it yourself, putting in 20 to 40 hours. From $2,500 to $8,000 with an independent consultant. From $10,000 to $40,000 with a research firm, where most of the price covers recruiting and paying respondents.
How many people do you need to interview?
Twelve qualitative interviews per customer segment are enough in most SMB markets: saturation, the point where you no longer hear anything new, arrives at around that number. For a quantitative survey, plan for 400 respondents for a margin of error of about 5 points, and 100 for a rough order of magnitude.
How long does it take?
From three to six weeks running it yourself at a few hours a week. From two to four weeks with a consultant. From six to ten weeks with a firm, about three of which go to recruiting respondents alone.
Can you do market research with no budget?
Yes, and it is the most common case. Statistics Canada, the Institut de la statistique du Québec, the Registraire des entreprises and Google Trends are free. A BAnQ library card, free for every Quebec resident, also opens paid commercial databases. The only line item that really costs money is data collection from a panel, and you can replace it with twelve interviews you conduct yourself.
What is the difference between market research and a business plan?
Market research checks that someone wants to buy. The business plan explains how you are going to sell it to them, and with what money. The research always comes first, and it feeds the plan's revenue projections. A business plan whose revenues are not backed by market research is a string of assumptions.
- Statistics Canada, Census Profile, Business Register and tables on business dynamics, accessed August 2026.
- Institut de la statistique du Québec, regional data and sector surveys (in French), accessed August 2026.
- Registraire des entreprises du Québec, enterprise register, accessed August 2026.
- Bibliothèque et Archives nationales du Québec, databases available with a library card, accessed August 2026.
- Données Québec, open data from the government and municipalities, accessed August 2026.
- Saturation threshold at twelve interviews: Guest, Bunce and Johnson, “How Many Interviews Are Enough?”, Field Methods, 2006.
- Price ranges of Quebec providers and project durations: Falia survey, August 2026.
Gabriel almost always takes your first call and carries out your audit. He builds the strategy starting from your growth goal: where to put your budget, which market to test and how to connect each lead to a real sale in your CRM. He mainly leads engagements for three goals: Optimize the profitability of your digital campaigns, Develop a new market, and Generate demand and growth. With Geneviève, he also works on organic search (SEO), AI visibility (GEO) and conversion rate optimization (CRO). The sales a Google Ads or Meta Ads campaign brings in depend on the page that receives the click. He writes mainly about marketing strategy, paid advertising and measurement.
About Falia →
SEO
Paid advertising
Conversion
AI visibility
Web design