Every leadership team walks into a quarter carrying beliefs about its competitors that nobody has checked: that your price sits at the bottom of the market, that the objection your reps hear most is the one the whole category hears, and that the feature beating you is the one on their homepage.
In our experience, at least one of those is wrong. Finding out means somebody has to go through the competitor’s sales process as a real buyer, and that is the work Mystery Demo does.
We book your competitors’ demos, take every call as a genuine buyer, record what the rep says and shows, and hand you the evidence on a Notion board. One competitor costs a flat 499 euros and takes one to two weeks.
So the real question is whether that evidence changes something you would otherwise get wrong. Interesting is cheap, and being wrong about a rival’s price for a quarter is not.
Measure the Artifact That Changed, Not a Modeled Lift
The temptation is to build a return calculation. Take a win rate, add a plausible lift, multiply by deal size, and produce a number with a decimal point in it.
That number is fiction, and everyone in the room knows it. Nobody can separate what one research project did to a win rate from everything else that moved in the same quarter.
So the honest unit of return is something smaller: a named internal artifact that changed, with an owner’s name on it and the date it changed.
A pricing rule, a battlecard field, the order of your demo, or the question that opens your next executive review. Evidence that confirms the rule you already have counts too, because it ends an argument your team would otherwise keep having.
You can walk into a room six weeks later and check whether that happened. A modeled win-rate lift cannot show what this project caused, which does not stop it arriving with two decimal places and a meeting.
Which Decisions Are Worth Testing This Way?
Three are worth testing first. They are illustrations rather than a ranking, because nobody can say in advance which of them pays best for you.
A pricing floor is only comparable once the conditions match, which is the whole of what a competitor pricing review captures. A headline number without its conditions attached is a rumor with a currency symbol on it.
.png)
Below is the pricing page of Mystery Demo’s public example board. The companies on it are invented, and the columns are the ones every real project fills in: each vendor’s pricing model, its floor, how openly the rep talked about that floor, and what it charges for, side by side.
A battlecard earns a rewrite when the line a competitor’s rep uses against you turns out to differ from the line your team assumed. In our experience, a rep rarely improvises objection handling, and the same line often comes back on that rep’s next call. That is why a battlecard rewritten from a recording beats one rewritten from a guess.
The third is a product question. A capability described as important and never shown live is a different fact from one demonstrated on screen, and that difference should change your next executive review. On the board, that difference gets its own column: what each vendor claims, next to what its demo showed and how it was shown.
The Limits of a Trained Observer
That distinction is where this method is strong, and it is worth being plain about where it stops. A mystery demo records what a competitor’s reps said and showed to one buyer across one evaluation. It is not a forecast of the competitor’s revenue, and it does not see inside their building.
What a trained observer does well is narrower and more useful: catching the moments of a sales call that a real buyer forgets by the next morning. The table below sets each decision beside the evidence that tests it and the claim that evidence cannot carry.
| The decision | Evidence to capture | What it does not prove |
|---|---|---|
| Your pricing floor | Metric, volume, minimum commitment, term, implementation charge, discount condition, quote validity | That their price predicts their revenue |
| A battlecard response | The exact objection line, the proof offered, the reframing question | That it is the line every rep uses |
| Executive review priorities | What was demonstrated live versus described, deferred, or promised | That an unshown feature does not exist |
One evaluation gets read by several people for different purposes: pricing reads the floor, sales enablement reads the objections, and product reads what was shown. One project, several decisions.
If one decision on your desk this quarter rests on what a competitor quotes, says, or shows, tell us about it on a short intro call. We will say which demos would settle it and what that would cost.
Is 499 Euros a Lot of Money?
What you are buying is one Notion board per project:
Operators who have sold B2B software take the calls, working from a buying scenario fitted to each competitor, so the rep gives a real evaluation instead of the generic tour.
It is a fixed published fee per competitor, which makes the comparison unusually simple to run. One competitor is 499 euros, five are 2,495, and ten are 4,990, and there is no second invoice: extra meetings, follow-up, and the analysis all sit inside the fee.
A single competitor takes one to two weeks, against four to eight for a full landscape. The fee does not change because the competitor is famous.
Set the Fee Beside the Decision
Every project starts by writing the decision down. Below is the objectives page of the same invented example board: the question the project exists to answer, and the objectives under it that every competitor is asked about.
From there we build the buying scenario, book each competitor’s demo through its normal front door, take every call, record it, and write the analysis against these questions.
Set the number against the decision, in that order:
A pricing rule that governs a large share of your outbound quotes is not a minor assumption to be carrying untested. Keep the fee beside that decision rather than inside an invented return model.
And if the decision at stake is genuinely smaller than the fee, you have your answer, and it is a perfectly good one.
A Finding With No Owner Changes Nothing
The failure mode here is not a bad piece of research. It is a good one that nobody acts on.
So the test at the end is one question, asked six weeks after delivery: did an artifact change, and can you name it?
| The artifact | Who usually owns it | What counts after six weeks |
|---|---|---|
| A pricing rule | Whoever signs off discounts, often pricing or revenue operations | The rule moved, or the record says why it stays |
| A battlecard field | Sales enablement | The field quotes the rival's recorded line and the answer to it |
| A demo path | The sales lead | The order changed because of what the calls showed |
| An executive review agenda | The founder or CEO | The question it opens with came from what was shown or only described |
If a pricing rule moved, a battlecard field got rewritten, or a demo path was reordered because of what the calls showed, and its owner can say which decision it improved, the project earned its keep. If the evidence confirmed the rule you already had, it earned it too, because the argument is settled on the record. If nothing moved and nothing was settled, it did not, whatever the quality of the analysis.
The recordings are the evidence, and the written analysis of what they add up to comes from the operator who was on the calls. That analysis is what turns a recording into a decision somebody can own.
Most teams already know which belief they would least like to be wrong about. Bring that one to a call, and Mystery Demo will tell you which competitors’ demos would settle it, then go and take them.
.png)
