Competitor pricing: most rivals do not publish a price, so the real work is making two different packages comparable
The seven public places a real number can be found and what each one proves, the normalisation step that turns two offers into one unit, a worked comparison where the cheapest option changes with headcount, the three ways a price list is built to resist comparison, and the record to keep so a number never travels without its source.
Published by AI Knows Us · Updated 20 October 2026
The search behind this is almost always the same question: what do they charge. The reason it is hard to answer is not secrecy. It is that what a rival publishes is a package, and what you need is a unit. A flat monthly fee for "up to 25 users" and a per-user price are not two prices, they are two shapes, and until you convert both to the same unit at your own customer's real size you do not have a comparison, you have two numbers next to each other. That conversion is most of the work, and it is the part that changes decisions.
This page covers where a real number can actually be found and what each source is evidence of, how to normalise two offers into one unit with the arithmetic done, the three devices that make a price list hard to read, what the finished comparison is allowed to change about your own pricing, the record to keep, and the two lines not to cross while collecting any of it.
Where a real number actually lives
Rank your sources by what they prove rather than by how easy they are to reach. A list price is not a paid price, and a paid price from one buyer is not a market rate.
| Source | What it proves | What it does not |
|---|---|---|
| A published pricing page | The list price and the inclusion list on a given date | What anybody actually pays after negotiation |
| The self-serve checkout or sign-up flow | The real transactable price, including the annual and monthly split | Enterprise or custom terms, which never appear there |
| Public contract awards and tender registers | A price somebody genuinely paid, with a date and a scope attached | That the same scope would be quoted to you |
| Rate cards and media kits | The opening ask, which is the number a negotiation starts from | The volume discount structure behind it |
| Marketplace and app store listings | A standardised price in a format built for comparison | Anything bought outside that marketplace |
| Partner, reseller and affiliate pages | The margin a rival is willing to give away, which bounds their cost | Their end customer price |
| Your own lost deals | What a real buyer was quoted, in your own market, this quarter | Anything general, because it is one data point from an interested party |
The last row is the one most businesses underuse and it is the strongest evidence available to you. A buyer who has just chosen somebody else will often say what they were quoted if you ask once, plainly, after the decision and without arguing with it. Write the answer down with the date and the deal size beside it.
The normalisation step, with the arithmetic done
Pick one unit and convert everything to it. The unit should be the thing your customer counts: a user, a location, a transaction, a thousand impressions, a tonne, a case, a month of access. Then build the comparison at three sizes, because one size hides the thing you most need to see.
Here is a worked example. The numbers are illustrative, chosen so the arithmetic is visible; use your own.
- You: 29 per user per month, everything included, no minimum.
- Rival A: 1,200 per month flat for up to 25 users, plus a one-off 2,400 onboarding fee, annual contract.
- Rival B: 19 per user per month, with storage and reporting in a 299 per month add-on that most buyers need.
The unit is cost per user per month in year one, with the one-off fee spread across twelve months.
| Headcount | You | Rival A | Rival B | Cheapest |
|---|---|---|---|---|
| 8 users | 29 | 175 | 56 | You |
| 25 users | 29 | 56 | 31 | You |
| 60 users | 29 | 43 | 24 | Rival B |
Rival A at 8 users is 1,200 plus 200 of spread onboarding, so 1,400 a month for 8 people, which is 175 each. At 60 users the flat fee buys a second block of 25, so 2,400 plus 200 is 2,600, which is 43 each. Rival B at 8 users is 152 plus the 299 add-on, so 451, which is 56 each.
The useful output is not the three rows, it is the crossover. You are 29 and Rival B is 19 plus a fixed 299, so the gap is 10 per user and the fixed fee is covered at 299 divided by 10, which is 29.9 users. From 30 users upward Rival B is cheaper than you, every time, and no amount of selling changes the arithmetic. That single number is worth more than a full competitor deck: it tells you exactly which deals you are structurally expensive in, and it is the input to a packaging decision rather than a discount.
The three devices that make a price list hard to read
These are ordinary commercial design rather than deception, and each has a specific counter.
- The unit swap. Per seat against per site against per transaction against per revenue share. Nothing compares until you have picked one unit and converted, and the rival whose unit is unlike yours is usually the one being compared to you incorrectly in the market.
- The bundle. A larger number that includes things you charge for separately, or a smaller number with the essentials unbundled. Counter it with an inclusion list: write the ten things a buyer in your market actually needs, then mark which are included at each quoted price. The comparison becomes a grid instead of a number.
- The threshold. Tiers that step at a user count, a volume or a feature. A customer sitting just under a step is getting a very different effective price from one just over it, so always run the comparison at a size just below and just above each threshold rather than at a round number.
Two further items belong on the sheet whenever they exist, because they move the real annual figure more than the headline does: a one-off setup, onboarding or migration fee, and a price that is only available annually. A monthly-equivalent figure that ignores either one is not a comparison.
What the finished comparison is allowed to change
A competitor number tells you about the market's reference point. It does not tell you what your own price should be, because your cost base, your capacity and the thing you are actually better at are not in their numbers. So put the levers in this order.
- The unit, first. If you are structurally expensive above a certain size, the answer is usually a different unit or a volume band at that size, not a discount on the old unit.
- The inclusion list, second. Moving one item in or out of the base package changes your effective price without touching the number, and it is reversible.
- The proof, third. If the comparison shows you more expensive and you intend to stay there, the page that holds the price has to carry the reason, stated as a specific difference rather than as a quality claim.
- The number, last. A price cut made from a competitor sheet rather than from your own margin arithmetic is the move most often regretted. The full decision, the reference points a price can legitimately be set from and how to change one without losing the base, is set out in this guide to setting a pricing strategy.
The record, so a number never travels without its source
One row per competitor per observation, and six columns. Anything missing a source and a date gets deleted rather than argued about, because an unsourced price repeated twice inside a company becomes a fact.
| Column | Why it is there |
|---|---|
| Date observed | Prices move, and a figure with no date cannot be trusted at the next review |
| Source, as a full address or a named document | So the next person can re-check it in one click instead of re-finding it |
| Unit as quoted | The raw shape, before your conversion, so the conversion can be audited |
| Number as quoted | Including the currency and whether tax is in it |
| What is included at that number | The inclusion list is the comparison |
| Who recorded it | So a questionable line has somebody to ask |
Set a refresh date on the sheet rather than updating it continuously. Quarterly is enough for most markets, and a self-serve category where a checkout page can change overnight deserves monthly. For what the finished thing looks like when every claim is traced to where it was found, see a competitor assessment with the source against every line.
Two lines not to cross
- Do not misrepresent who you are to get a quote. Posing as a buyer, using a false company, or signing up under a pretext to reach a rival's pricing page is a reputational and contractual risk that buys you a number you could usually have obtained from a public source or a lost deal.
- Never discuss or coordinate prices with a competitor. Collecting public information about what a rival charges is ordinary commercial research. Agreeing what to charge, or exchanging future pricing intentions, is price fixing and is unlawful in most markets regardless of how informal the conversation was. The difference is observation against agreement, and it is not a fine line.
Questions about competitor pricing
How do I find out what a competitor charges when they do not publish it?
Work down the evidence list rather than guessing. Try the self-serve checkout, public contract awards and tender registers, marketplace listings, published rate cards, and partner or reseller margin pages. Then ask the buyers you lost what they were quoted, which is the strongest source available to you because it is your market, this quarter, at a real deal size.
What is a competitive pricing strategy?
Setting your price with the market's reference point as the main input: matching a rival, pricing deliberately above with a stated reason, or pricing below and accepting the volume that has to follow. All three only work once you have normalised the comparison to one unit, because matching a package you have not converted usually means matching a number that is not comparable to yours.
How often should I refresh the sheet?
Quarterly for most markets, monthly where a checkout page can change overnight. Put the refresh date in the document, because the failure mode is not an out of date sheet, it is an out of date sheet that nobody knows is out of date.
Should I match a cheaper competitor?
Run the crossover first. If they are cheaper only above a certain size, the problem is your unit at that size and the fix is a volume band or a different unit rather than a general cut. If they are cheaper everywhere and you intend to hold your price, the inclusion list and the stated reason have to do the work, and they have to be on the page where the price is.
Is it legal to collect competitor pricing?
Collecting information a competitor has made public is ordinary research and is how most markets work. The two things to avoid are obtaining it by misrepresenting who you are, and discussing or agreeing prices with a competitor, which is a different activity entirely and is unlawful in most markets.
How many competitors should the comparison cover?
Three is usually enough to see the shape: the one buyers mention most, the cheapest credible option, and the most expensive credible option. Adding a fourth rarely changes a decision, and a list of ten is how this exercise turns into a spreadsheet nobody reads.
The price a buyer hears before they reach your page
There is now a stage before the comparison you are building, and you are not in the room for it. A buyer asks ChatGPT, Gemini, Perplexity, Claude or Google's AI Overviews who the options are in your category and what they cost, and gets back two or three names with figures attached. Those figures are assembled from whatever the web says, so they can be an old list price, a tier you retired, or a number from a review site, and they are quoted with no date on them. Three things reduce that: a pricing page that states the unit and the inclusion list in text rather than in an image, a dated last-updated line on it, and retired prices removed rather than left on an old landing page. Ask your own category question in two assistants, keep your own name out of the prompt so the answer comes back clean, and write down the number it gives for you. AI Knows Us runs those buying questions across the main assistants and records which companies get named and what each answer was built from.