How to do a competitor analysis (with template)

A competitor analysis records what your competitors sell, at what price, with which message and through which channels, and what that means for your own marketing. "Competitive analysis" means the same thing in day-to-day marketing. A market and competitive analysis adds the market itself: size, trends, customer segments.

Every analysis is a snapshot. Competitor monitoring is the ongoing part: the same competitors and the same questions, checked again every week or every month. Prices, ads and taglines change without notice, so last year's analysis describes last year's market.

This guide covers both. You'll get the method in seven steps, a template to download and a worked example.

What the sources say

The statements about sources and frameworks come from Meta's and Google's own documentation and from Porter's own articles. Every link is listed under "Sources" below.

Public sources cover most of the work. Websites, pricing pages, job ads, reviews and social channels cost nothing, and anyone on the team can check them. You won't see what happens behind a login or in sales calls; a good analysis says so instead of filling the gap with guesses.
The Meta Ad Library shows ads running in the EU. Since August 2023, Meta has archived all ads that target people in the EU, with the dates they ran, targeting parameters such as age, gender and location, and information on who was served the ad. Meta says these ads stay in the public library for a year.
The Google Ads Transparency Center shows Google ads. Google describes it as a searchable repository of advertisers and the ads they've served on Search, Display, Gmail and YouTube, filterable by date and targeted location. For ads in the EEA, Google also discloses targeting information and the total number of recipients per ad.
Rankings can be observed; traffic can only be estimated. Record observed rankings with the query, country, device and date. Label third-party traffic estimates separately. Search Console data is available only for properties you can access, meaning your own or ones an owner has granted you access to, which rules out competitors' sites in almost every case.
Porter's Five Forces describe an industry, not a competitor. Porter introduced the model in Harvard Business Review in 1979 and revised it in 2008. The five forces are customers, suppliers, potential entrants, substitute products and established rivals; together, he writes, they define an industry's structure. That makes it a poor checklist for competitor X. It's useful earlier, when you choose whom to analyse, because Porter warns that managers often define competition too narrowly, "as if it occurred only among today's direct competitors."
Language models guess when they don't know. A 2025 paper by Kalai and colleagues describes how language models produce plausible but incorrect statements instead of admitting uncertainty. For a competitor analysis, that means every price and every metric from an AI tool needs a source someone on your team has opened.

How to run it

The method works for five competitors as well as for fifteen; the effort grows with every name on the list. Start with the template so every finding lands in the right place from day one.

Competitive analysis template to copy

One row per competitor, one column per question. The last two columns matter most: without a source nobody can check a finding, and without a date nobody knows in three months whether the price still holds. The blank template with five rows is a CSV file to download that opens in Excel or Google Sheets. Or paste this header row and split it at the semicolons (Data, then "Text to Columns" in Excel or "Split text to columns" in Sheets):

Competitor;Audience;Offer;Price;Positioning/message;Channels;Strengths;Weaknesses;Source;Date

What goes in each column:

Competitor: name and website
Audience: who the website addresses, ideally in their own words
Offer: the products or services that compete directly with yours
Price: list price and model (subscription, package, "contact us")
Positioning/message: the homepage tagline, verbatim
Channels: where they advertise and publish, including live ads
Strengths: what reviews praise and what the offer does better than yours
Weaknesses: recurring complaints and gaps in the offer
Source: one URL per finding
Date: the day you checked the finding

Competitor analysis example (hypothetical)

Say you run marketing for a coffee roaster that ships beans on subscription. One of five competitors is "North Roasters". The company is made up, and so is every entry in this row; it only shows what a filled-in template looks like.

Competitor: North Roasters, northroasters.example
Audience: people working from home ("for everyone who works from home")
Offer: coffee subscription with three blends, grinders as add-ons
Price: €16.90 per 500 g on subscription, €18.90 without
Positioning/message: "Freshly roasted, at your door in 48 hours"
Channels: Instagram, newsletter, eight active ads in the Meta Ad Library, all for the sample pack
Strengths: delivery time as a clear promise; many reviews praise shipping
Weaknesses: several reviews mention broken aroma valves; no offer for offices
Source: homepage, pricing page, Ad Library, review profile (one URL each)
Date: 1 September

So what? North Roasters sells speed to home workers and doesn't address business customers at all. If your sales team confirms that offices ask about coffee subscriptions, test that gap with a landing page of your own. The table gives you the hypothesis; only the test tells you whether it holds.

Step by step

Write down the question. Put in one sentence which decision the analysis should inform, such as a price change, a pitch or a campaign. Without that question you'll collect everything and decide nothing.
Choose the competitors. Take the direct competitors with the same offer for the same audience, add the indirect ones who solve the same problem differently, and include whoever ranks at the top for your most important search terms. Start with five to eight competitors if research time is limited.
Read the website and pricing page. Fill in audience, offer, price and the homepage tagline, the tagline verbatim. Save a screenshot or PDF, because the page may look different tomorrow.
Check the ads. Look up each competitor in the Meta Ad Library and the Google Ads Transparency Center. Note which offers they promote and which message keeps coming back. Record how long an ad has run; the library does not show whether it is profitable.
Read job ads, reviews and social channels. Job ads show where a company is investing: a new market, a new product, an in-house performance team. Reviews give you the weaknesses in customers' own words. On social media, which topics get a reaction tells you more than the follower count.
Check search visibility. See which of your key terms a competitor outranks you on. Record observed rankings with the query, country, device and date. Label third-party traffic estimates separately. The SEO & GEO Researcher adds visibility in AI answers to that picture.
Assess and decide. Record the facts about each competitor in the template first. Then do a separate SWOT for your own company: your strengths and weaknesses, and the opportunities and threats that follow from what you found about the competitors. Then build a feature and price matrix, with the features customers compare when they buy as rows and every vendor as a column, yours included. Close with half a page: what you saw, what it means, and what you'll do now.

How to measure it

An analysis goes stale the day a competitor changes their pricing page. So decide up front who updates it and how often. Weekly pays off in markets that run heavily on paid social and promotions; monthly is enough where offers and prices move slowly. Repeat the full analysis with SWOT and matrix less often, for example before annual planning, before a launch, or when a new player shows up.

Where AI helps and where it doesn't

AI takes the gathering off your hands: reading pages, summarising ads, spotting what changed since last time. That's the part that eats hours and that rarely has an owner.

It can't check or decide for you. When a model can't find a price, it may invent a plausible one. So ask for a link behind every claim, and open it for anything that feeds a decision. Whether you respond to a competitor's price move is still your team's call.

Hand the recurring part to AI coworkers

On Sokosumi, competitor monitoring runs as a standing brief. You name the competitors, the moves you care about (pricing, launches, paid social, hiring, messaging) and the rhythm. Four coworkers gather from public sources: Meta Ads Library reads the paid social creative, Website Traffic Analysis estimates traffic, sources, engagement and geography for several domains in one run, News Research reads recent English-language coverage with headline-level sentiment, and Hannah runs a full competitive analysis when you want to look at one competitor in more depth.

What comes back is a finished document with what changed since the last run: positioning and messaging changes quoted verbatim, new campaigns, traffic and channel shifts, pricing and packaging moves linked to the page they came from, and a short "so what" for your own roadmap. You can add your own domain to the list so your site sits in the same table. Anything behind a login, or said in a private sales conversation, is out of scope.

For a single company there's the Company Researcher. It researches one company per run from public sources and returns a report on its market, competitors and positioning with source citations; check the evidence for each claim you use (30 credits per run as of September 2026). To compare several companies, run it once per company and put the reports side by side.

An account is free. Each run costs credits. Costs depend on the coworkers and number of runs; check the displayed credit price before starting. Plans and the credits included per seat are on the pricing page.

What to check at every update

Price and packaging changes on pricing pages, with the old and new value.
New or ended ads in the Meta Ad Library and the Google Ads Transparency Center, and the message they carry.
Changed taglines on the homepage and landing pages, verbatim.
New job ads that point to a new market or product.
Rankings for your most important search terms, with query, country, device and date.
One line on what you'll do and who does it. Without that line, the update was reading, not work.

Where teams get this wrong

Looking only at direct competitors. Customers also compare you with substitutes: a freelancer, a different product category, or "we'll do it ourselves."
Entering findings without a source and a date. Three months later nobody can tell whether a price is still right, or ever was.
Filling in the Five Forces as a competitor checklist. The model explains how an industry works and where the profit goes; it says nothing about what competitor X will advertise next week.
Copying numbers from an AI chat without checking them, especially revenue, market share and prices.
Reporting third-party traffic estimates as measured values, or recording rankings without the query, country, device and date.
Doing the analysis once and never opening it again.

Sources

Competitor analysis: common questions

What should a competitor analysis include?

Per competitor: audience, offer, price, positioning and message, channels, strengths and weaknesses, plus the source and the check date for every finding. It ends with what that means for your own offer and what you'll do next.

What's the difference between competitor analysis and competitor monitoring?

A competitor analysis (or competitive analysis) is a snapshot at one point in time. A market and competitive analysis adds market size, trends and customer segments. Competitor monitoring is the ongoing update, weekly or monthly.

How often should you do a competitor analysis?

Do the full analysis with SWOT and a feature and price matrix before big decisions: annual planning, a launch, a price change, a new player in the market. In between, monitor weekly if your market runs on paid social and promotions, monthly otherwise.

What tools can I use for competitor analysis?

The Meta Ad Library and the Google Ads Transparency Center are free and open to anyone, and so are competitors' websites, job ads and review profiles. SEO tools show observed rankings for other domains and estimate their traffic; record the query, country, device and date, and label traffic estimates separately. Search Console won't help here, since it only covers properties you can access. On Sokosumi, AI coworkers do the recurring gathering and return a report with sources.

Is Porter's Five Forces useful for competitor analysis?

For choosing whom to analyse, yes. Porter's five forces (customers, suppliers, new entrants, substitutes, established rivals) show who besides your direct competitors puts pressure on your prices. For comparing individual competitors, the template and a feature and price matrix are the better tools; the SWOT comes afterwards, for your own company.

Can AI do a competitor analysis?

The research half, yes: reading pages, summarising ads, spotting changes and writing it all up with sources. You still have to check it, because language models invent numbers when they can't find them. And what your team does with the result isn't something a report decides.