For founders
Or you are about to hire ahead of revenue, or launch the second product, or open the next market, and the honest answer to what happens is that nobody knows. Counterfirm turns the guess into a distribution: four hundred runs of your actual company, a band around the answer, and a list of which assumption the whole thing is resting on.
Runs on your own WordPress install. No API key needed for the simulation itself.
The advice you get is confident in inverse proportion to how much of your data the person has seen.
Everyone has a view on your pricing. Almost nobody has looked at your cohort behaviour, your contract lengths, your gross margin or the share of your revenue sitting with three logos. A model that has all of those and still gives you a wide band is more useful than a person who has none of them and gives you a number.
The four that actually matter early
The highest leverage thing most early companies can do and the one they defer longest. The sweep runs twenty five price points at eighty replications each and gives you the curve, including how flat it is near the top, which is usually the reassuring part.
Raise priceCash out now, productivity in a few months, quota later still. The run carries ramp, onboarding months and the attrition that takes one of them back out again before they have paid for themselves.
Hire a teamIt takes attention off the first one. That is a real cost and it almost never appears in the plan. The run puts the cannibalisation, the attention and the new revenue in the same twelve months.
Launch a productEntry cost now against share later, against an incumbent who already lives there. The band on this one is wide, and the width is the finding.
Enter a new marketRunway
Runway calculated as cash divided by burn is a single path through a world where nothing varies. In a run, churn draws differently every time, new logos draw differently, collections and hiring both move, and what you get back is the share of runs in which cash crosses the floor and the month the first crossing happens.
That number changes how you raise. "We have eleven months" and "one run in five has us under the floor by month seven" are different conversations with an investor, and the second one is the one you can act on.
Thin data
The twin gets built anyway, on industry starting points, and the readiness score tells you how much of the answer is standing on them rather than on you. That is not a consolation prize. Knowing that seventy percent of your pricing answer comes from a default elasticity is itself a finding, and it tells you the single document worth digging out.
Bands are wide when the record is thin, and they narrow as you upload. Nothing here pretends to more precision than your files support, which is the opposite of the failure mode of a spreadsheet with four decimal places in it.
For the board meeting
What is the downside, what has to be true, and how would we know early. A brief answers those in that order by construction: the p10 is the downside, the ledger is what has to be true, and the tripwires are how you would know.
It also removes the worst dynamic in a board meeting, which is two people with different mental models arguing about a conclusion. Put the assumption on screen, change it to the number your investor believes, re-run, and see whether the conclusion survives. Usually it does, which ends the argument faster than defending it would.
Honestly
It will not tell you whether the idea is good. Product market fit is not in the equations. The engine models a company that sells something people already buy, and it will happily simulate twelve months of a business whose real problem is that nobody wants the product.
It will not help much at pre revenue. With no customers, no churn history and no price, almost every number is a default and the twin is closer to a tutorial than a model of you. Be honest with yourself about which one you are looking at.
It will not find your next channel, write your positioning or tell you which competitor is about to raise. It has no data about the world outside your folder.
Practicalities
No. Explore is 149 a month, Operator 390, Boardroom 890, and Enterprise is a conversation. Yearly is ten months for the price of twelve. All of that may change before general release, and it is said here so nobody is surprised later.
The product is new and there is nothing to show here yet. There are no testimonials on this site, no logo wall and no usage numbers, because inventing them would undermine the only thing the software is actually for, which is being straight about what is known and what is guessed.
In your own WordPress install, in a directory outside the media library that is not served over the web. They are parsed into chunks so numbers can be cited back to a line, and the file and its chunks are deleted together when you delete the source.
Yes. The worked example is Harborline Components, an invented contract manufacturer with 14.24 million dollars of revenue, twenty four customers, a largest customer at eighteen percent and a price it has not moved since 2023. It is invented, it is labelled as invented, and every scenario is already set up on it.
It is the decision with the most leverage and the least evidence behind it. A sweep of twenty five price points is two thousand simulations and comes back as a curve with a best point on it.