Competitive Analysis Framework for a Week, Not a Quarter

A competitive analysis framework built for a slide gets presented once and quoted for a year. The version that changes a week is smaller and harder to fake. It asks four questions, in order, and it cannot recommend anything until each answer has a source.

If your current process starts with a template and ends with a color-coded grid, you already own the artifacts. You are missing the loop. A grid that was true in March is a liability in October, because people will quote it. The job of a weekly framework is to make last month’s quote embarrassing as soon as the page it came from changes.

What a competitive analysis framework is actually for

A competitive analysis framework is a decision filter. It exists to answer one question for one rival: did anything this period change what we charge, what we ship, or what we say on the next call? If the answer is no, the output is the word “no,” written down, with a date. Teams that cannot record “no” rebuild the deck every quarter to feel busy.

This is not a competitor research template. That template is how you learn a company in an afternoon when they are new to the set: eight fields, each with a place to look. Fill it once when a rival earns a slot. Do not refill all eight fields every Monday. You will skip weeks, and then the template is stale in a way nobody notices.

It is also not a competitor matrix. The matrix is a weighted scorecard and a map. It is the right tool when you are arguing where to stand, or when a board wants a picture. Weighting twelve criteria every week teaches you nothing, because the weights should not move that often. Run the matrix when the weekly loop has produced a real shift, not as a substitute for the loop.

And it is not a SWOT grid. A competitor SWOT analysis earns its place when every cell has a claim, a source, and an implication. That is a thinking tool for a decision you already suspect matters. It is a bad weekly habit, because four boxes invite adjectives. Adjectives do not survive a sales call.

Use those three when the question is big. Use the weekly framework so you know when the question got big.

Four questions, in an order you cannot skip

Ask them in this order. Skipping ahead is how a team writes the recommendation first and then goes looking for a quote that supports it.

1. Who is this about, this week? One company. Not the market. If you cannot name them, you are doing landscape work, and landscape work has its own process. The who includes the buyer they are selling to, not just the logo. A pricing change aimed at a different buyer is a different event from the same percentage change aimed at the buyer you share.

2. What changed, in their words? One diff. The homepage sentence, the tier name, the review complaint, the job post, the paragraph on the changelog. Paste the before and the after, or write “no dated before, so this is a snapshot, not a change.” A snapshot is allowed. Pretending a snapshot is a change is not.

3. So what, for a buyer we actually meet? One sentence a prospect could say out loud. “Their Business plan now includes SSO, so security review is no longer a reason to wait for a contract.” If you cannot say it the way a buyer would, you do not have a so-what. You have a summary of their website.

4. What do we do, or not do, by Friday? One of three moves. Update the talk track. Change our page or our package. Do nothing, and say why. “Monitor the situation” is not a move. It is question 2 with a longer deadline. Put a date on the nothing: revisit on a specific Monday, or kill the note.

Leave “what is their strategy?” off the weekly pass. Strategy is what you infer after three diffs point the same way. One diff gets the diff, not a paragraph.

The evidence rule that keeps it honest

Every answer carries a source and a date, or it does not go in the note. The source is a URL you can open, a review you can point at, or a fetch from a pricing directory with the fetch date. A colleague’s memory of a demo is a lead, not a source. Demos are edited.

The money line is where the rule pays for itself. The largest number on a pricing page is often not what a buyer pays. As of our directory’s fetch on 2026-09-29, Linear’s pricing shows Basic at $10 per user per month and Business at $16, with the yearly toggle active and no separate monthly price printed. Free is $0 (unlimited members, 2 teams, 250 issues), there is no free trial, Enterprise is contact sales on annual billing, and Business is the featured tier. “Linear is $10,” without the toggle and the date, is how a wrong sentence survives into a deal.

Two bans. Do not import a price from a roundup you did not check. Do not write a percentage unless both numbers come from fetches. A slider, a team-size picker, or a yearly toggle will invent a change the company did not make. If the two captures could be two positions of the same control, write “possible measurement,” and stop.

Competitive positioning is what you defend once question 4 says the rival’s claim hits yours. The weekly note is the tripwire. Do not rewrite positioning off a hero test that reverted six days later. Question 2 should have said “test.”

How Monday actually goes

Block thirty minutes. Not a workshop. Thirty minutes, same time, same doc.

Spend the first ten on the surfaces that can take a deal from you: the pricing page, the homepage hero, and the changelog or the release note. You are looking for a diff against what you stored last time. No stored baseline means this week is a snapshot week. Say so at the top of the note. The next Monday is the first real one.

Spend the next ten on one other surface, and rotate it: reviews, then job posts, then a comparison page about you. Watching every surface every week is how the half hour becomes an afternoon you cancel.

Spend the last ten on question 3 and question 4 for the single diff that matters. If nothing matters, write the date and “no decision” next to each tier-one name you checked. That line is the product. It is what stops a teammate from making a decision on a rumor later in the week, because the rumor now has to beat a note that says you looked.

Open the matrix or the SWOT once a quarter, and only if the weekly notes changed a weight or a cell. If they did not, leave the chart alone.

Competitor benchmarking is question 2 across the set: the same facts, dated, without a fresh narrative each time. The Monday note is the narrative, and it should be short enough to read aloud. A preamble means it is already too long.

Who this is for, and who should not bother

A founder or a single product marketer with three to seven rivals can run this without an analyst. The constraint is the archive. If last month’s sentence is gone, this week is fiction. A shared doc with dated pastes beats a system nobody opens. A platform earns its place when the diff is already there with the date on it, and the paste is the step you keep skipping.

Skip it if you have one rival and you already talk to their customers every week. You are not short of information. You are short of a place to put the decision. Question 4 is the whole document, and the rest is overhead.

The one move worth making this week

Pick one rival. Write the four questions on a single page. Answer question 2 with a paste from their pricing page or their homepage, including the date you looked, and write “snapshot” if you do not have last month’s version. Then force question 4 into one line: talk track, page, package, or nothing, with a revisit date.

If you cannot finish in thirty minutes, the competitive analysis framework is already too big. Cut a field before you add a category. Next Monday you run it against the note you just wrote. The second pass is the one that can surprise you.

Outmano is an AI-powered competitive intelligence platform: pricing, SEO, content, roadmap, and review monitoring, with analysis on each change, delivered in a dashboard, alerts, a weekly digest, and over MCP to your own AI. The Monday note still has to be yours. The diffs should not be. outmano.com

Frequently Asked Questions

What is a competitive analysis framework?

A competitive analysis framework is a repeatable set of questions that turns public evidence about a rival into a decision: change the talk track, change your page or package, or do nothing. The useful ones are short enough to run every week. A framework that needs a quarter to fill is a research project, and research projects do not catch a pricing page that moved on a Tuesday.

How is this different from a competitor matrix?

A matrix scores several rivals on weighted criteria so you can see who is ahead and where the map is empty. It is a picture, and pictures go stale quietly. The weekly pass watches one company for one change and forces a decision. Rebuild the picture only when those notes say the picture changed.

What counts as evidence in a weekly competitive note?

A line you can reopen: the page’s own words, a review, a job post, or a dated fetch of a pricing page. The date sits next to the line. A summary written from memory, a number from a listicle you did not check, and a percentage you computed off a slider are not evidence. If two captures could be two settings of the same control, label it as a possible measurement and leave it out of the decision.

How long should weekly competitive analysis take?

Thirty minutes for a small set, most of it on the pricing page, the homepage, and one rotating surface. The first week is often a snapshot, because you do not have a baseline yet, and it can run long while you paste. After that, the time goes to the diff and the decision. If it regularly takes more than an hour, you are researching, not running the framework.

When should you redo the whole competitive analysis?

When three weekly notes point the same way, or when you are about to change your own price, your homepage claim, or the segment you sell to. Those are the weeks for the matrix, the SWOT, or a fresh pass on a rival who just entered the set. A calendar reminder that says “refresh competitive analysis” with no trigger produces a new deck and the same decisions.