How to Analyze Competitor Positioning: 10 Sources and What Buyers Hear

Almost every guide on this subject sends you to the competitor’s homepage, has you fill in a two-by-two grid, and stops. That produces a record of their marketing copy, which is the one part of their positioning they wrote specifically for you to read. It is worth having and it is not an analysis.

Positioning is not a claim, it is a place in somebody’s head. So there are two datasets here, and they come from different places. One is what the company publishes about itself, which is fully controlled and completely available. The other is what the market repeats back, which is not controlled at all and takes real work to collect. The sources below are split across both, and the reliability column reflects which side each one evidences.

1. 1Name the decision the analysis has to serve. Whether to change a headline, whether to contest a category, or how to answer one objection are three different jobs. A positioning study with no decision attached becomes a slide nobody acts on, and it dates within a quarter.
2. 2Extract their claimed position into five components. Read their homepage, product pages, pricing page and boilerplate, and write down the alternatives they name, the attributes they claim are unique, the value they promise, who they say it is for, and the market category they file themselves under.
3. 3Collect the received position from sources they do not control. Review text, third-party comparison pages, community threads and your own lost-deal notes. Take verbatim wording only. The aim is the sentence a buyer uses when nobody from their marketing team is in the room.
4. 4Write the gap as one sentence. The output is of the form: they claim X, buyers repeat Y, and the difference is Z. A gap that cannot be stated in a sentence is usually two findings tangled together, or an opinion without evidence behind it.
5. 5Date every change from archived pages. Pull their homepage and product pages at roughly six-month intervals and note when the category word, the headline promise and the proof points changed. Repositioning is expensive, so each change is a decision with a date on it.
6. 6Draw the map using axes taken from buyers. Pick the two attributes that appear most often in the received wording, not the two your team finds most flattering. Then place both companies on the claimed map and the received map, because they rarely land in the same place.
7. 7Convert the gap into one messaging change with an owner. A positioning gap is only worth finding if something changes: a headline, a discovery question, an objection response or a proof point. Record the change, the owner and the date, and re-check it against the next quarter's lost deals.

This is the organising idea of the page. Sort every source by whether the competitor controls it, and the pile stops being a list of tabs to open and becomes two columns you can compare. The left column is an intention. The right column is a result. Companies are judged on the right one.

Free-text notes about a competitor’s positioning are impossible to compare, either against another competitor or against the same company six months later. Extracting into fixed fields fixes that, and the cleanest decomposition available is the five components April Dunford sets out in Obviously Awesome. It was written for positioning your own product, and it works just as well pointed at somebody else’s, because each component is separately observable.

Fill those five fields for each competitor and one thing tends to jump out immediately: which component is doing the work. A company whose category is unremarkable but whose target customer is razor-sharp is running a segment strategy. A company with a broad target and a single loud attribute is running a feature strategy, and feature strategies are the ones that get copied. The conclusion belongs in a differentiation matrix alongside the rest of your set, because a position is only meaningful relative to the others.

1. Their homepage and category words

Copy the headline, the subheading and the boilerplate paragraph from the footer or the press page into your notes verbatim. Do not summarise, because the summarising is where you accidentally improve their copy. Then isolate the category word, the single noun they use to say what kind of thing they are. That word is the most expensive decision on the page: it determines which comparisons they invite, which search demand they can capture, and which budget line they get paid from. Note whether the word is an established category, a modifier on one, or an invention.

2. Archived versions of their positioning pages

Pull the homepage at roughly six-month intervals for the last two or three years and line up the headline, the category word, the named logos and the top pricing tier. What you get is a dated timeline of positioning decisions, which no competitor publishes and few of them could reconstruct themselves. The same archive technique applied to their customer page answers a different question and is covered under competitor customers.

3. Their own comparison and alternatives pages

A vendor comparison page is advertising, and the verdicts are worthless. The list of names is not. Building and maintaining a comparison page costs real effort, so a company only builds one for rivals it meets often enough to justify the work. Read the page for two things: who is on it, which tells you their real competitive set, and which attribute rows they chose, which tells you the ground they believe they win on. Note anyone conspicuously absent, since the biggest name in a category is often left off deliberately.

4. Review-site categories and grid placement

Two different things live here, and the second is the one people misread. The categories a product sits in, and which of them carries the bulk of its reviews, tell you where the market files them rather than where they file themselves. The grid position is a different animal: on G2 the vertical axis is Market Presence, which blends review counts with company-size and web-visibility signals from third-party data. A high placement is therefore partly a measure of how big and how visible a vendor is, not only of how good it is. Read it as reach, and read the horizontal satisfaction axis as reception. The review text itself deserves its own method, set out under competitor reviews.

5. Analyst evaluations and vendor reprints

Evaluations of this kind are participatory: vendors respond to a request for information, brief the analyst, demonstrate the product and supply customer references. That matters twice over. It means the write-up partly reflects how the vendor chose to present itself, which is a positioning artefact in its own right. And it means the cautions or weaknesses paragraph is close to unique: a named third party stating in writing what a competitor is not good at, which no other public source will give you. Vendors licence reprints and host them ungated, so the paragraph is usually free to read even when the report is not.

6. Their pricing and packaging page

Positioning claims are cheap and packaging decisions are not, so where the two disagree, believe the packaging. The entry price and any seat minimum define the smallest customer they will accept. The tier names carry segment language they have committed to. What sits behind “contact sales” marks the point where they want a human in the deal, which is a statement about deal size. The full method for reading and sourcing those numbers is under competitor pricing.

7. Product marketing job adverts

This is the earliest positioning signal available and almost nobody reads it. A product marketing job description has to describe the work honestly enough to attract the right candidate, which means it often names the category the company intends to move into, the segment it is trying to reach and the launch it is staffing for. A posting for a product marketing manager for a segment they do not currently name on their site is a repositioning being resourced. The general reading method for adverts is under competitor hiring.

8. Paid search and social ad copy

Ad copy is the fastest-moving positioning surface a company has, because it is rewritten against a conversion rate rather than against a brand guideline. Public ad libraries let you read active creatives and their run dates without an account, so you can see which promise they keep and which they quietly retire. The single most useful comparison is ad copy against the homepage: when the ads have been saying something different for months, the ads are usually right about what sells and the website is preserving what the company wishes were true. Where the libraries are and what each one exposes is covered under competitor ad spend.

9. Win/loss interviews and closed-lost notes

This is the highest-value source on the page and the one that costs the most to work, because somebody has to ask buyers questions and write down what they said. It is also the only source that gives you the received position from people who had budget. Two rules make it usable: record the buyer’s exact words rather than the rep’s interpretation, and ask what they thought each vendor was for, rather than which features they compared. Our win/loss interview questions include the framing that gets positioning language out of a buyer rather than a feature list.

10. Conference talks, podcasts and investor calls

Executives explain their position differently when the audience is peers, analysts or shareholders rather than prospects. The vocabulary loosens, the strategy behind the copy gets stated out loud, and the reasoning appears. Search for the founder or the chief executive by name across podcast episodes and conference sessions, and read any earnings transcript if the company is listed. This is also where you catch a repositioning before the website changes, because leaders talk about a new direction for months before marketing rewrites the homepage.

A change in wording is not automatically a repositioning, and treating every campaign as a strategy shift is how competitive teams manufacture false alarms. Four markers separate the two, and they hold in almost every software category.

• The category word changed. They now describe themselves as a different kind of thing. This is the strongest single marker because it resets which comparisons they invite and which buyers they reach.
• The named proof changed. Different logos, different case study industries, different outcome metrics. Proof follows the segment they are actually selling to, so it moves when the segment moves.
• The entry tier changed. A raised or removed starting price, a new seat floor, or a self-serve plan appearing or vanishing. Packaging is the most expensive thing on this list to change and the least likely to be cosmetic.
• The change held. New wording that survives two consecutive archive snapshots and reaches the pricing page, the boilerplate and the sales collateral is a position. Wording that appears and disappears within a quarter was a test.

Two of the four is enough to record a dated repositioning event. Once you have a few of those for each company, the sequence tells you more than any single snapshot: a competitor that has moved upmarket twice in three years is telling you where the next move goes, and it is usually further in the same direction. Keep the timeline beside the rest of the set in a competitive landscape rather than in a document per company, because the interesting pattern is who moved after whom.

Search results for this topic are full of numbered frameworks, and a good share of them do not exist outside the pages selling them. Knowing which is which saves you from building an analysis on a list somebody invented to rank for a query.

Real and attributable

• Positioning itself. Jack Trout used the term in a 1969 article, and Al Ries and Jack Trout published Positioning: The Battle for Your Mind in 1981. The founding claim, that a position is something occupied in the buyer’s mind rather than something asserted in copy, is exactly why the received half of this page exists.
• Functional, symbolic and experiential. Park, Jaworski and MacInnis, Journal of Marketing, 1986. A three-way split of brand concepts, and the closest thing to an academic answer to the “types of positioning” question. Note it is three, not four.
• Reverse, breakaway and stealth positioning. Youngme Moon, Harvard Business Review, May 2005. Genuinely useful for reading an unusual competitor: reverse positioning strips away attributes the category treats as compulsory while adding unexpected ones, breakaway positioning attaches the product to a different category entirely, and stealth positioning conceals the category a product really belongs to.
• Porter’s generic strategies. Competitive Strategy, 1980. Three strategies, with focus splitting in two, which is why they are usually drawn as four. Useful as a consistency check rather than a label.
• Porter’s four corners. Future goals, assumptions, current strategy, capabilities. This is what “Porter’s framework for competitor analysis” refers to. Do not confuse it with the five forces, which is a model of an industry’s structure and says nothing about any individual company in it.

Not canonical, whatever the search results say

The “four types of positioning” and the “5 P’s of positioning” have no attributable origin, and the versions in circulation contradict one another. The five Ps query is usually reaching for the marketing mix, which is McCarthy’s four Ps from 1960, extended to seven for services by Booms and Bitner, and which describes an offer rather than a position. The “five criteria for understanding competitors” does not exist either. Saying so is more useful than inventing a fifth item to fill a heading, and it is the honest answer to the question the searcher actually asked.

1. 1Quote it, do not summarise it. Every claimed-side finding should be traceable to a sentence you can paste, with the URL and the date you read it. Summaries silently smooth contradictions out of the source material.
2. 2Require three buyers, not one. A single reviewer describing them a certain way is an anecdote. The same description from three unconnected buyers is the received position, and the threshold matters more than the wording.
3. 3Check the packaging agrees. If the claimed position says enterprise and the entry tier is thirty dollars a seat with a credit card form, the position is aspirational. Packaging is the cheapest lie detector on the page.
4. 4Date every field. Positioning findings rot invisibly, because the page still looks current long after the company has moved. An undated note is unusable within two quarters.
5. 5Test it against a real deal. Take the gap sentence into the next competitive call and see whether the buyer recognises it. This is the only verification step that involves someone with money, and it settles arguments no amount of desk research will.

This is one of the safest areas of competitive research, because a position is worthless unless it is published. Marketing sites, pricing pages, ad libraries, review platforms and conference talks are all deliberately public. The boundaries that do exist are mostly about how you gather buyer language and what you then say in public.

• Say who you are when you interview a buyer. Win/loss conversations are the richest source here, and the rule the profession settled on is that a buyer should know which company is asking before they start answering. Someone who knows gives you quotable material; someone who works it out afterwards gives you a complaint.
• Do not publish a comparison you cannot source. If you turn this analysis into a public comparison page, every factual claim about the competitor needs a link to their own material and a date. Marketing that misdescribes a rival is a straightforward legal problem in most markets, and an easily avoidable one.
• Do not post or solicit reviews you did not earn. Reviewing a competitor as a pretend customer, or asking staff to do it, breaches every major platform’s terms and is treated as a deceptive practice by consumer regulators. It also corrupts the one dataset this method depends on.
• Use their trademarks descriptively, not decoratively. Naming a competitor to compare honestly is normal and lawful. Reproducing their logo and colours so your page resembles theirs is a different matter, and the risk sits in the presentation rather than in the name.

• Why they chose it. The research, arguments and internal politics behind a position are never published. Proxy: the sequence of changes over three years, which shows the direction of travel even when the reasoning is invisible.
• Whether it is working commercially. No outside source tells you what a position converts at. Proxy: your own contested win rate against them, which measures exactly the thing you care about and is first-party.
• Their internal positioning document. The one their sales team is trained on differs from the website, often substantially. Proxy: what buyers report being told in demos, which is that document as delivered.
• The next repositioning. Nobody announces one in advance. Proxy: product marketing job adverts and executive interviews, both of which describe the intended position months before the site does.
• How a position is received in a market you do not sell in. Reception is local, and your deal notes only cover your own territories. Proxy: review text filtered by reviewer country, plus their local-language sites, which are frequently positioned differently from the English one.

Run the full extraction once or twice a year, because positions are designed to be slow. Between those, keep a short monthly check on the four markers that indicate a real change: the category word, the headline promise, the named logos and the entry tier. That check takes minutes per competitor once you know where to look, and it is the difference between noticing a repositioning in week two and reading about it in a customer’s renewal conversation.

Add the received half continuously rather than in a batch, because it never arrives all at once. The discipline that makes this work is unglamorous: after every competitive loss, paste the buyer’s own description of the winning vendor into the same file, with a date. Four quarters of that is a better dataset than any positioning study you could commission, and it accumulates whether or not anybody schedules the work.

The decay here is unusually quiet. A repositioning does not arrive as an announcement; it arrives as a homepage that reads slightly differently one Tuesday, and the old version stops existing the moment the new one deploys. Unless somebody happened to look that week, the change is undated forever, and a dated timeline is most of the value of this work. Meanwhile the received half keeps moving on its own as new reviews land and new deals close.

Checking a dozen quiet marketing pages every week to catch four words is not work a person keeps up, which is what competitive monitoring platforms exist to carry. Flares follows competitor messaging, category claims, packaging and proof points, and reports what changed and when, which is the dated record nobody can reconstruct after the fact. What it cannot do is the received half. No platform can tell you what a buyer thought a competitor was for, because that sentence only exists in a conversation somebody has to go and have.