A Competitive Analysis Framework for Canadian Businesses That Leads to Decisions
Most competitive analysis is a folder of screenshots that nobody opens twice. A useful one answers a short list of questions and ends with decisions.
Why most competitive analysis fails
Ask a leadership team who their competitors are and you will usually get the names of businesses that look like them: same size, same services, same part of town. Ask the same team why they lost their last three deals and the answers are often different. The client went with a national firm. The client hired someone in-house. The client decided to wait a year.
That gap is where most competitive analysis goes wrong. It studies the companies the team thinks about rather than the alternatives the buyer considered. It collects everything (logos, colour palettes, follower counts, website screenshots) and concludes nothing. And it is done once, filed, and never updated.
The framework below is designed to avoid all three. It is deliberately light, because a competitive review that takes three months is out of date by the time it lands.
Step one: choose the right competitors
Build the list from three directions, and expect them to disagree.
- Business competitors. The organizations you actually lose work to. Ask sales, check the CRM’s lost-deal notes, and ask a few recent clients who else they talked to. These are the ones that matter most.
- Search competitors. Whoever shows up when buyers search for what you do, in Google, in the map results and in AI answers. These are often directories, national chains, review sites or a business you have never considered a rival. They are competing for your buyer’s attention whether you count them or not.
- The alternative. Doing it in-house, doing nothing, using a cheaper tool, or sticking with the current supplier. In many B2B categories, the status quo is the competitor you lose to most often.
Keep the final list short: three to six names you will study properly, plus a note on the alternative. A dozen shallow profiles are less useful than four deep ones.
Step two: see what buyers see
Review each competitor the way a buyer encounters them, in roughly the order they do. You are not grading design; you are recording what a buyer would conclude.
Search, maps and AI answers
Search for the services you sell, with and without a place name, on a phone, from a couple of locations. Note who appears in the map results and who appears below them. Then ask ChatGPT, Gemini or Perplexity the questions a buyer would ask (who should I hire for this in Halifax, what is the best option for that in New Brunswick) and record who gets named and how they are described. Our guide to how ChatGPT recommends businesses explains what shapes those answers.
The website
Read the home page and the main service pages. What do they claim? Who do they say they are for? What proof do they show: named clients, case studies, credentials, guarantees? How easy is it to take the next step, and what is the next step? Request a quote or book a consultation if it is reasonable to do so; how a competitor handles an inquiry tells you more than their website does.
Reviews
Read competitors’ Google reviews, and those on any platform your buyers use, for themes rather than star ratings. What do customers praise them for? What do the unhappy ones complain about? Recurring complaints about a competitor are one of the clearest gaps you will find.
Advertising
Google’s Ads Transparency Center, Meta’s Ad Library and LinkedIn’s ad library show the ads an advertiser is currently running, publicly. You will not see spend or results, but you will see their messages, offers and how long they have kept them running. An ad that has run unchanged for months is probably working for them.
Signals of direction
Job postings, new locations, new service pages, press releases and leadership changes show where a competitor is heading, which is often more useful than where they are.
Step three: write it down in one comparison
For each competitor, and for your own business, record the same handful of things in plain language:
- Claim: the main thing they say about themselves, in their words.
- Audience: who they appear to be built for.
- Proof: what evidence backs the claim up.
- Offer: what a buyer can actually do or get, and how pricing is presented (published, quoted, packaged, hidden).
- Visibility: where they show up in search, maps, AI answers and ads.
- Reputation: the recurring themes in their reviews, good and bad.
Now read down the claim column. In most markets, three or four businesses are making nearly the same claim (quality, service, experience, local) with nearly the same proof. That is what buyers see too, which is why they default to price or proximity. Doing this honestly for your own row is the hardest part, and the most valuable.
Step four: find the gaps
Gaps come in a few recognizable shapes.
- An unclaimed position. Something buyers care about that nobody credibly owns: speed, a specialty, a type of client, bilingual service, transparent process.
- A claim without proof. Everyone says it; nobody shows it. The first to prove it owns it.
- A recurring complaint. The thing customers keep criticizing competitors for is a promise you might be able to keep.
- A visibility gap. Searches, places or questions where no strong competitor appears, or where the answer is a directory rather than a business.
- A neglected audience. A segment, region or language market the established players are underserving.
Not every gap is worth filling. Some are empty because nobody wants them. Test each against two questions: do buyers actually value this, and can we credibly deliver it? Turning a gap into a position is a separate piece of work, covered in our brand positioning framework for competitive markets.
Step five: decide, then set a date to look again
The analysis is only finished when it has changed something. End with two or three decisions, each with an owner. Typical examples: rewrite the main service page to lead with the claim nobody else can make; add the proof that supports it; stop competing on a search term where the economics do not work; open a French-language version of the services a competitor is ignoring; fix the inquiry process because a competitor answers in an hour and you answer in two days.
Then put a lighter review in the calendar. Ads, search results and AI answers change often enough that a quarterly look at those, and a full refresh once a year or before any major decision, is a sensible rhythm.
A worked example
Take an illustrative case: an accounting firm with three offices in New Brunswick. The partners name two local firms as their competitors. Sales notes show they actually lose most often to a national firm on larger clients and to cloud bookkeeping tools on smaller ones. Search results for their services are dominated by the national firm and by directories. AI assistants, asked to recommend an accountant for a growing business in the province, name the national firm and one of the local competitors.
The comparison shows all four firms claiming personal service and experience. None shows clear proof of either. Reviews of the national firm repeatedly mention staff turnover and not knowing who is handling the file. Nobody is speaking directly to owner-managed businesses going through a transition, which is where this firm does its best work.
The decisions: build the service pages and proof around that specialty, make continuity (one named partner per client) an explicit promise, and publish a French version of the key pages for the firm’s francophone clients. Three decisions, each testable, each traceable to evidence. That is a competitive analysis doing its job.
Doing it yourself, or not
Most of this can be done in-house with a few focused days, provided someone is willing to be honest about the business’s own row. The usual reason to bring in an outside view is exactly that: a team rarely sees its own claims the way a buyer does. A competitive review is part of every marketing strategy engagement Brand Butter runs, and if you are not sure whether you need an analysis, a strategy or an audit, marketing audit vs marketing strategy will help you decide. For a free outside look at how you appear in search and AI answers, start with the brand visibility report.
Key Takeaways
- Study the competitors you actually lose to, the ones that outrank you in search and AI answers, and the alternative of doing nothing.
- Review each competitor as a buyer would: search and maps, AI assistants, website, reviews, public ad libraries and signals of direction.
- Record claim, audience, proof, offer, visibility and reputation in one comparison, including an honest row for your own business.
- Look for unclaimed positions, unproven claims, recurring complaints, visibility gaps and neglected audiences, then test each against buyer value and your ability to deliver.
- End with two or three owned decisions and a calendar date for the next review.
Frequently Asked Questions
How do I find out what ads my competitors are running?
Google’s Ads Transparency Center, Meta’s Ad Library and LinkedIn’s ad library publicly show the ads an advertiser is currently running. They do not reveal spend or results, but they show messages, offers and how long an ad has been live, which is often enough to understand a competitor’s pitch.
How often should a business do a competitive analysis?
A full review once a year, or before a major decision such as a rebrand, a new website or a new market, is usually enough. Ads, search results and AI answers move faster, so a quick quarterly check of those keeps the picture current without much effort.
Who should I count as a competitor?
Start with the organizations you actually lose work to, then add whoever appears when buyers search for your services, including directories and national players. Include the option of doing nothing or doing it in-house, which in many categories is the most common alternative.
What should a competitive analysis produce?
A short comparison of each competitor’s claim, proof, offer, visibility and reputation, a list of the gaps worth pursuing, and two or three decisions with owners. If it ends with screenshots and no decisions, it was research rather than analysis.

