Competitive profile matrix: AG1 vs Grüns, scored

What a competitive profile matrix is

A competitive profile matrix, or CPM, scores you and your rivals on the handful of factors that decide success in a market. Each factor gets a weight for how much it matters, the weights add up to 1.0, and each company gets a rating from 1 to 4 on each factor. Weight times rating, summed, gives one score per company.

It is best known from Fred R. David's Strategic Management textbooks, and it turns up in strategy courses and pitch decks for the same reason: one number per company is easy to present. That single number is also the tool's biggest weakness, and this article shows why.

How it works

Critical success factors

The few things a company must get right to win in the market, as the buyer sees it. Not everything you do well. Six to ten factors is plenty.

Weights that sum to 1.0

Each factor gets a share of the total. A factor at 0.25 counts two and a half times as much as one at 0.10. The weight belongs to the market, not the company, so every company in the matrix is scored with the same weights.

Ratings from 1 to 4

The textbook scale: 4 is a major strength, 3 a minor strength, 2 a minor weakness, 1 a major weakness. Each company gets its own rating on each factor.

Weighted scores and a total

Weight times rating gives the weighted score for each cell. The totals run from 1.0, a major weakness on everything, to 4.0, a major strength on everything. 2.5 is the middle.

The usual complaint about a CPM is that the ratings are opinions. The fix is to tie every rating to a number you can measure, with the rule written down before you look at who scores what.

Worked on AG1 and Grüns

AG1 and Grüns both sell daily greens supplements, and both were measured on the same public sources. Every number below was measured on 28 September 2026. The Loop against Hears benchmark shows the same measures on a live page.

The factors are chosen for someone deciding which greens supplement to buy: what other customers say, whether the brand shows up when they ask, and whether the site works. Each factor has one measured number behind it.

The rating rule

With only two brands, quartiles do not work: there is nothing to rank between. So each factor gets fixed bands, set in advance. Where someone outside the two brands already draws a line, the band uses it. Google calls a site responsive at 200 milliseconds or less (web.dev), so that is where a 4 starts. The other bands are mine.

Read against those bands, on 28 September 2026:

• Customer rating: AG1 4.3, Grüns 4.6. AG1 rates 3, Grüns 4.
• Recent experience: AG1 3.41, Grüns 3.40 on the newest reviews. Both rate 1.
• Presence in AI answers: AG1 named in 56% of answers across five assistants, Grüns in 18%. AG1 rates 4, Grüns 2.
• Search authority: domain rating 75 against 65. AG1 4, Grüns 3.
• Content depth: 396 blog posts indexed against 28. AG1 4, Grüns 2.
• Owned audience: 684,434 Instagram followers against 166,317. AG1 4, Grüns 3.
• Site responsiveness: 248 ms against 183 ms. AG1 3, Grüns 4.

The weights

The weights are my judgement, not a measurement. I weighted them for a greens buyer: what customers say counts most (0.25 overall, 0.20 recent), being named when the buyer asks an assistant next (0.15), and the four remaining factors at 0.10 each.

AG1 scores 3.05, Grüns 2.70. Both sit above the 2.5 midpoint. AG1 wins on the factors that take years to build: authority, content, audience and AI answers. Grüns wins on the two a buyer feels directly: the overall rating and the speed of the site.

Note the recent experience row. Both brands score 1, because both sit at 3.4 on their newest reviews, well below their overall ratings. The row adds 0.20 to both totals and separates nothing. It still belongs in the matrix: a factor where both brands are weak is a finding, even when it moves no totals.

Move one weight and watch the result

Take 0.15 away from presence in AI answers and give it to customer rating. Nothing is measured differently. Only my opinion of what matters has changed.

The winner flips. The break-even sits at a shift of about 0.12: move more than that from AI answers to customer rating and Grüns comes out ahead.

A lead of 0.35 in the total was built on one weighting call. That is not a flaw in this example. It is how every CPM works, and most of them do not show it. If you present one, present the weight that would change the answer, and let the room argue about that instead of the total.

How to make one

1. Name the buyer whose decision you are modelling. The factors for a first-time greens buyer differ from those for a retailer choosing a supplier.
2. Pick six to ten critical success factors, each tied to one number you can measure for every company.
3. Write the rating rule for each factor before you look at the numbers. Quartiles if you have four or more companies, fixed bands if you have fewer.
4. Set the weights so they sum to 1.0, and label them as judgement.
5. Measure every company on the same day from the same sources, and date the matrix.
6. Multiply and sum. Weight times rating per cell, then a total per company.
7. Test the weights. Find the smallest shift that changes the order and show it next to the total.
8. Read the rows, not only the total. The rows tell you what to do. The total only tells you who is ahead under one set of opinions.

What it cannot tell you

A CPM compresses. AG1 has 4.1 times Grüns' Instagram audience and a domain rating 10 points higher, and both gaps come out as the same single step on the rating scale. If you need the width of each difference, a gap analysis keeps it.

The factors are only what can be measured. Price, taste and the formula itself are missing. The report has Grüns' store prices but no comparable price for AG1, so price is left out rather than guessed. For a greens buyer that is probably the biggest hole in this matrix.

It says nothing about sales, or about why anyone buys. A lead on authority and content does not show that AG1 sells more, or that those factors are the reason anyone chooses it.

And the result is only as neutral as the weights. The bands are written down and the numbers are measured. The weights are opinion, and the total inherits all of it.

FAQ

What is a good score on a competitive profile matrix?

With ratings from 1 to 4 and weights summing to 1.0, totals run from 1.0 to 4.0, and 2.5 is the middle. Above 2.5 means more strengths than weaknesses on the factors you chose. The score only means something next to the other companies in the same matrix.

What is the difference between a CPM and an EFE or IFE matrix?

The EFE and IFE matrices score one company on its external opportunities and threats, or its internal strengths and weaknesses. A CPM scores several companies side by side on the same factors, mixing internal and external ones.

How do you choose the weights?

From the buyer's decision, not from what your company is good at. Customer interviews, review themes and search questions are better inputs than a workshop vote. Whatever you use, label the weights as judgement and show the one that flips the result.

How is a CPM different from a strategy canvas?

A strategy canvas draws each company's level on each factor as a line and leaves the shape to speak. A CPM weights the factors and adds them into a single score. The canvas shows where companies differ. The matrix claims who is ahead.