Competitive Analysis Frameworks: A Practical Guide (with Real Examples)

Most competitive analysis you read about — SWOT, Porter's Five Forces, BCG matrix — comes from strategic planning textbooks written in the 1980s for industries that move at quarterly cadence. They're designed to produce decks, not decisions.

The frameworks that actually work in 2026 share three traits:

1. They're comparative. A useful framework forces you to read your competitor's claim against your own claim, in the same axes. Frameworks that analyze "the market" instead of "you vs. them" produce summary, not signal.
2. They're concrete. Every claim must cite a specific page, quote, or move. Frameworks that allow you to write "they have strong brand awareness" produce slop. Frameworks that force you to write "their homepage hero says X, ours says Y, here's the gap" produce decisions.
3. They're action-ranked. A framework that produces 30 observations is worse than one that produces 5 ranked moves. The whole point is doing something different.

Three frameworks meet these criteria reliably for B2B competitive analysis: Dunford's 5-step positioning canvas, Jobs-to-Be-Done, and Ries + Trout category ownership.

April Dunford's framework (from Obviously Awesome) is the cleanest tool for positioning analysis. It asks five questions of any product:

1. Alternatives. What would the buyer use if your product didn't exist? (Not "competitors" — alternatives, which includes "do nothing" and "build it themselves.")
2. Unique attributes. What does your product do that the alternatives can't? (Features, not benefits.)
3. Value. What do those attributes enable the buyer to do that they couldn't before? (Benefits, not features.)
4. Best-fit customer. Who cares most about that value? (The buyer segment that wins from the unique attributes.)
5. Market category. What space does the buyer mentally place you in? (Often different from how you describe yourself.)

To run a competitive analysis: fill out the canvas for the competitor and for your own product, side by side. Where the boxes differ is where positioning differentiation lives. Where they're identical is where the buyer can't tell you apart — and you have a problem.

Concrete example. Linear's homepage in early 2026 said "The product development system for teams and agents." Their Dunford canvas read: Alternatives = Jira, Asana, in-house tooling; Unique attributes = AI-native workflows, agent integration; Value = AI agents do work alongside humans; Customer = modern engineering teams; Category = "agent platform" (explicitly NOT "issue tracker").

Compare that to Jira: Alternatives = Linear, Asana; Unique attributes = enterprise governance, regulatory compliance; Value = predictability + auditability; Customer = enterprise engineering teams; Category = issue tracker.

These are positioned for different buyers despite sharing surface-level functionality. Read against each other, the differentiation jumps off the page.

Christensen's Jobs-to-Be-Done framework asks: what job does the buyer hire this product to do? Phrased: "When I [situation], I want to [motivation], so I can [outcome]."

For competitive analysis, JTBD reveals when two products that look similar are actually competing for different jobs — and when two products that look different are competing for the same job.

Example. Notion and Linear look like different products. But for a 5-person engineering team, both can be hired for the same job: "When my team needs to coordinate work across people and time zones, I want one tool that holds the plan and the work product, so I can avoid the cost of context-switching between tools." For that team, Notion and Linear are alternatives — even though they're sold into different categories.

For the same team's 50-person enterprise version, the job shifts: "When the company needs to track release readiness across 12 squads, I want a tool that enforces process and produces compliance reports, so I can ship on schedule without audit risk." Now Notion is out and Linear is competing against Jira.

The job is more stable than the product category. Competitive analysis grounded in JTBD catches competitors who don't look like competitors on the surface — and that's where the most strategic threats come from.

Al Ries and Jack Trout's Positioning introduced the idea that buyers don't have unlimited mental space — they remember 1-3 brands per category, and the brand that "owns the word" wins. Salesforce owns "CRM." HubSpot owns "inbound." Slack owns "team chat."

For competitive analysis, this framework asks: which word does the buyer associate with each competitor? If you can't answer that question for your own product in one word, you have a positioning problem — your buyer can't either.

When competitors rewrite their homepages, they're usually attempting to occupy a new word. Linear declaring "issue tracking is dead" was a deliberate move to leave the "issue tracker" category and claim "agent platform" instead. Notion adding "AI workspace" to their hero was an attempt to expand from "wiki/docs" into "AI-native productivity."

These category-occupation attempts succeed or fail based on whether the buyer ratifies them. The buyer ratifies them when (a) the product genuinely delivers on the category claim, and (b) no incumbent already owns the word.

In competitive analysis, watch for category-occupation attempts — they signal where the competitor thinks the future of the market is. Even if their attempt fails, they're telling you what they believe.

Combining the three frameworks, a productive competitive analysis takes ~90 minutes per competitor:

0-15 min. Read their homepage, pricing, product page, and careers page. Take notes on the literal copy — quote it.

15-30 min. Fill out the Dunford canvas. Be specific. "Strong customer support" is not an attribute — "24/7 phone support with sub-3-min response SLA" is.

30-45 min. Write their JTBD as one sentence: "When [situation], I [motivation], so I can [outcome]." Then write yours. If they're the same, the differentiation has to come from another axis.

45-60 min. Name the category word they're occupying. Name yours. Are they trying to occupy a new one (check the homepage hero — is it different from 6 months ago)?

60-75 min. List 3-5 concrete moves you could make. Each should be a single sentence, verb-led, with a clear effort/impact rating.

75-90 min. Pick one and start it Monday.

The output isn't a deck — it's a decision. If your analysis didn't produce a Monday-morning action, you over-analyzed.