← All posts

You Iterate Your MVP. Now Iterate Your Business Plan.

You iterate your MVP because the feedback is cheap and repeatable. A user churns, you learn something, you change the product, and the next user never knows there was an earlier version. Now look at how founders treat the business plan. They write it once and then iterate it the most expensive way available: by pitching it to live investors and trying to read the room. I argue you should iterate the plan the way you iterate the product, against cheap feedback, before you ever spend an investor on it. The point is simple: iterate the plan before you pitch it.

I have been on both sides of this. I started Books.com in 1992, the first online bookstore, back when "buy a book on the internet" was the pitch nobody believed. Later, I built Flashline, a software component company that was acquired by Oracle. Then I ran an accelerator, Flashstarts, which meant I read a large stack of business plans and sat across the table from the founders who wrote them. The founders who got funded were rarely the ones with the best first draft. They were the ones whose plan had been through the most versions before it reached the right investor.

The plan deserves iteration, too

Lean startup gave us a habit and a vocabulary. Build, measure, learn. Ship the smallest thing, watch what real users do, change the product, ship again. Founders internalized this so completely that it now feels obvious. Of course, you iterate the product. Iterating is the job.

Then those same founders write a business plan as if it were a legal filing. One draft, polished for a week, and from then on, the only thing that ever tests it is a pitch. The assumptions inside it (who the buyer is, what they will pay, how you reach them, why now) are exactly the kind of claims lean startup taught us to test cheaply and often. With the plan, we stopped. We took the document holding our riskiest assumptions and agreed to iterate it only in the one room where mistakes cost the most. That is the problem this piece is trying to solve.

Every investor is a bridge you cross once

Here is the thing: founders underrate investors. Investors are a finite, non-renewable resource. For your stage and sector, there might be 30 or 40 funds that realistically make sense, and that number is illustrative but closer to the truth than most founders would like to admit. Each one is a single shot. They talk to each other, they remember your name, and a no on a weak draft from your top-choice fund is a door that usually does not reopen this round.

So when you iterate your plan by pitching it, every loop costs you one of your best leads. Fifty pitches and fifty rejections are not fifty data points. They are fifty crossed bridges and a plan that learned too slowly to matter, because by the time the story is finally good, the investors who would have loved it have already passed. You used them up teaching yourself what was wrong.

That is the trap. The only critic founders have had is also the one critic they cannot afford to spend on a draft.

Iterate against a critic, not your cap table

We built a managed agent, Alex, that reads a plan the way an investor would in the first ten minutes and answers in minutes rather than months. You submit the plan. It comes back with a structured read of where the story is strong, where it only asserts, and the one weak assumption most likely to sink the raise. It also returns a rough valuation range. Then you fix the weakest part and submit again. You can iterate as many times as the work needs and burn nothing, because the critic is not on your cap table and does not talk to the funds on your list.

The valuation number is the part people grab first, so let us be precise about what it is and is not. It is wide on purpose, directional, and not investment advice or a number to put in a term sheet. Its job is to be a feedback signal you can move. When a sharper "why now" or a real piece of traction nudges the range up, you have learned something concrete about which part of your story carries weight, and you learned it without spending an investor to find out. That is the point: improve the plan with cheap feedback, then walk into the room with the version that already survived the hard questions.

The first read is free, because the point is to get founders into the loop, not to gate it. We are a small company running this as an MVP ourselves, with limited capacity and a first-come, first-served approach. We would rather be honest about that than pretend at a scale we do not have yet.

Current view, subject to change

I want to label the opinion as an opinion. The claim here is that iterating the plan against a cheap, honest critic before you pitch gets founders funded with fewer, better meetings, rather than grinding through a long list of nos. I believe that, and I am running the experiment in public. What would change my mind is a stretch of founders who iterate diligently, watch their read improve, and still raise no faster than founders who walk into every meeting cold. If the iteration does not change outcomes, the thesis is wrong, and I will say so. So far, the founders who treat the read as a loop, not a grade, are the ones who arrive ready.

Final thoughts

You would never ship version one of your product to your entire market and learn from the wreckage. You ship it small, read the signal, fix the weakest thing, and iterate until it is ready. Do not treat your plan any differently. It holds your riskiest assumptions, and it deserves the most iteration, not the least. Now you can give it that iteration without spending a single investor.

Iterate the plan until it is as good as you can make it. Then send it.

You can put a plan through its first read, free, at coworkers.global/intake.

Regards,

Charles Stack
Founder, Coworkers.Global

Research by Alex, a managed startup-advisor agent at Coworkers.Global.

The valuation range described here is directional and intended as feedback on a business plan. It is not a valuation opinion, an appraisal, or investment advice.

Coworkers.Global is an AI staffing agency. We place managed agents into organizations that need dedicated expert knowledge work. A managed agent is an AI specialist provisioned for a specific role, trained on your context, supervised by a person, and accountable for its output. The first, Alex, evaluates startup business plans for fundability, informed by human expertise and research, and calibrated against real investor decisions. We are early-stage and pre-revenue, so we lead with the quality of our judgment rather than customer logos we don't yet have. Your managed AI coworker.
Get new posts by email

What we're learning building a startup with managed agents, plus notes on raising. Monthly newsletter, no spam.