How do you start charging for work you've been doing for free?

You start by accepting that the client is not being unreasonable. Once free work has run long enough, it becomes the terms of the relationship in their mind, and any correction is felt as a loss. Time the change to a natural boundary, name the trade rather than the increase, and price forward instead of backward.

Why does the price-increase email never land the way the articles promise?

Because the articles are answering a question about wording, and the problem is not the wording.

Search for help with this and you will get a hundred versions of the same page: how to tell a client you are raising your rates, what to say, when to send it, how to frame the value. Every one of them assumes the number was right once and now needs updating. That assumption is what breaks. In the situation most founders are actually in, the number was never wrong. The work moved, quietly, one accommodation at a time, and each accommodation taught the client something about what the relationship includes.

Picture an agency running an account that started at eight hundred dollars a month of ad spend. It now runs thirty thousand across two platforms. The fee has never moved. Somewhere in year one the operator also built a CRM and a set of automations, and has been maintaining them ever since without billing a dollar for it. He knows exactly what has happened. He can describe it precisely and a little sheepishly.

Now he wants to correct it. He is looking for the right words.

There are no right words, and I would rather tell him that than sell him a script, because a script that fails in this situation makes everything worse. He is about to send a message that says, in effect, this thing you have had for a year now costs money. To him that is a return to fair terms. To the client it is a withdrawal.

What does the research say about why it lands as a betrayal?

That people judge a change against whatever they have come to expect, and that "eliminating a gain" and "imposing a loss" are not the same thing to them, even when the dollars are identical.

Daniel Kahneman, Jack Knetsch and Richard Thaler ran a set of telephone surveys in Toronto and Vancouver in the mid-eighties, asking randomly selected residents to judge whether various business decisions were fair. The results were published as Fairness as a Constraint on Profit Seeking in the American Economic Review, and one pair of questions is the whole of your problem in two paragraphs.

Question A. A car dealer has been selling a popular model at list price. He now sells it at $200 above list. 71% of respondents judged that unfair.

Question B. A car dealer has been selling that same model at a $200 discount below list. He now sells it only at list price. 42% judged that unfair. 58% said it was acceptable.

Same two hundred dollars out of the customer's pocket. The only difference is whether the concession had hardened into the reference point. In the first case the list price gives everyone an unambiguous baseline and the increase is coded as a loss against it. In the second the baseline is blurry, and the change can be read either as a loss or as the end of a gift.

Your situation is worse than either. You have been giving the work away for a year, so for your client the discounted arrangement is the list price now. There is no other baseline in their head. And the paper anticipates exactly how that happens: the authors note that a practice which begins as an exception "may spread slowly until it evolves into a new norm — and is no longer unfair." That sentence was written in 1986 and it is a better description of scope creep than anything the project management field has produced since.

The same researchers measured the underlying force four years later. In their endowment effect experiments, people who already held a mug demanded a median $5.75 to give it up, while people buying the identical mug would pay $2.25. Across markets and repeated trading rounds the seller-to-buyer ratio held around 2 to 2.6 times. Their conclusion was that undertrading comes from a reluctance to part with entitlements.

So you and your client are doing different arithmetic. You are asking them to add a line worth, say, five hundred a month. They are being asked to give up something they hold, and holding it makes it worth roughly double to them. A well-written email does not touch that.

Is this just scope creep, and isn't everyone told to avoid it?

Everyone is told, and it keeps getting worse.

The Project Management Institute's 2018 Pulse of the Profession, based on responses from 5,402 professionals, found that 52% of projects experienced scope creep, up from 43% five years earlier. The spread inside that number is the useful part: 33% at the highest-performing organisations against 69% at the worst. Roughly a two-to-one difference in the same economy, on the same kinds of work.

That gap tells you scope creep follows from whether a boundary was written down and whether anyone was watching it. Weather does not produce a two-to-one spread.

But notice what the advice built on that finding is for. Change orders, defined scope documents, approval workflows: all of it is prevention, aimed at the next engagement. It is genuinely good advice and it is useless to the founder standing in the middle of a live relationship carrying two years of absorbed favours with no contract end in sight. I have written before about charging for out-of-scope work when the request arrives, and about why raising prices on clients you already have is usually the wrong move. Both of those assume you catch it at the right moment. This piece is about what to do when you didn't.

Is your client actually the problem here?

Almost certainly not, and I want to say that clearly because the research above can be read as a warning about your client's character. It isn't one.

In twenty-odd years of doing this I have worked with very few takers. Most business owners are not sitting there calculating how much more they can extract. The clients of mine who have received the most free work are friends, and what they actually say is how much do I owe you and don't forget to invoice us, because they are generous people who do not want to be the person taking advantage. My own church, and churches have a reputation here, came to me about marketing help and said show us what this is really going to cost.

So the fairness research is telling you why you expect to be punished, and why that fear runs well ahead of the risk. The reference-point effect is real, and it is the reason the conversation feels like a betrayal from where you are standing. It is rarely the reason the conversation goes badly, because most of the time it doesn't.

Here is the part I would rather not write. I built a website for someone years ago and I am still doing the updates six years later. There are one or two like that. I have an arrangement now where I charge a little, and it is honestly not the same as what the work is worth. It is hardest when they are people you know and like, and I have had to get somewhat cold about it, because you have to talk about money. There is no version of running a business where you get out of that. If you are doing something for free, you are the one who has to say so, out loud, and being uncomfortable is not a reason to skip it. That is one of the genuinely hard parts of owning the thing, and you put on your big boy pants and do it.

The other thing I would change if I were starting again is the shape of the generosity itself. I have moved away from giving work away and toward a revenue share: I will help you get this off the ground and I get paid when you are making money. Some of those never make money and I end up working for nothing, and when that happens it is on me, because I did not manage to make them successful. That is a different failure from being taken advantage of, and it feels different, because the expectation was never that it was free.

And for what it's worth, the recurring problem in my own work has been revisions rather than pure favours. A thing gets signed off, and then it comes back, and then it comes back again. Same accretion, different door.

What actually works instead of a price-increase email?

Four moves, in rough order of how well they work.

Price the next thing rather than the last one. Everything already absorbed is sunk, and trying to reprice it is where the betrayal lives, because it requires the client to agree that they have been getting something for nothing, which is a claim about their character as much as about your invoice. Draw the line at today and never bill backwards. The next new request gets a number before the work happens, said plainly and without apology: that one sits outside what we set up, and it runs about this much. You are declining to extend the arrangement, which is a smaller and much easier thing than correcting it.

Trade instead of adding. The cleanest correction is usually a swap rather than more money. The retainer holds and something comes off the plate, or the new thing goes on and an old thing comes off. This works because it leaves the reference transaction intact; the client's total stays where they expect it and only the contents change. It also forces a conversation about what they actually value, which is generally worth more to you than the fee you were going to ask for.

Attach the change to a boundary that already exists. A contract renewal, a fiscal year, a platform migration, a scope review you put in the calendar months ahead. The reason the car dealer question splits 71 to 42 is that a clear reference point makes a change feel arbitrary, and a boundary hands you a new reference point that has nothing to do with you deciding to charge more. If your engagement has no boundaries at all, creating one is this quarter's job, because you cannot reprice something that never ends.

Then label the next favour as a favour, out loud, at the time. This is the prevention half and it is nearly free. Kahneman and his colleagues observed that firms prefer temporary discounts to permanent price cuts precisely because a discount can be withdrawn without violating the reference price, and the same logic works at your scale. It costs one sentence: I'll take care of that one, it's outside scope but it's a small thing. Say that, and the favour stays a favour. Say nothing, and the favour becomes a term.

And where you want to be generous, consider changing the shape of it rather than the price. A revenue share, or payment that starts when the thing you built starts working, keeps the expectation honest from day one. You may still end up doing a lot of work for nothing. But nobody ever believed it was free, which means there is no correction to make later and no reference point to violate.

That last one is the single highest-return habit in this entire piece and it takes eleven words.

Where does this sit in the architecture?

In Lifetime Value, which is where most people are surprised to find it, because it feels like a pricing problem and pricing feels like sales.

The reason it lives there is what unbilled scope does to a relationship over time. A founder who has been absorbing work for two years has gone quietly resentful about that client, and resentment shows up in delivery long before it shows up in a conversation. Response times stretch. The proactive ideas stop. The account gets the minimum because the founder has unconsciously decided it does not deserve more, and the client feels the temperature drop without knowing why.

Then that client leaves, and the founder tells himself they were price-sensitive.

They were over-served, under-charged, and slowly neglected by someone who could not say what was wrong. That is an offer architecture failure with an emotional delivery mechanism, and it is one of the most reliable ways a good account turns into a departed one. The same clients you are quietly resenting are the ones your next introduction was supposed to come from.

The framework does not hand you a template for this. Every business needs its own version, because what counts as inside the boundary in a fractional finance engagement looks nothing like what counts in a landscaping contract. You adapt it to your work. Nobody can adopt one from a book, and the ones sold in books are usually change-order forms with a philosophy stapled on.

Putting it to work

Take an hour with your three largest accounts.

1. Write down what they are actually receiving. Not what the agreement says. Every recurring thing you do for them, including the ones you would describe as "it's nothing." Most founders find between four and eight items that were never priced.

2. Mark each one with the date it started. You are looking for how long each favour has been running, because duration is what turned it into an expectation. Anything over about six months is settled in their mind, whatever your contract says.

3. Put a number beside each, for yourself only. Hours a month, at your real rate. Add it up. This number is for your own calibration and it will probably annoy you. Sit with that rather than acting on it today.

4. Find the next boundary on the calendar. Renewal, year end, a project completion, a platform change. If there isn't one within six months, your first job is to put a review date into the arrangement so one exists next time. Repricing comes after that.

5. Draft the trade. For each account, write the swap you would propose at that boundary: this comes on, that comes off, or the number moves and here is what changes with it. A trade is a conversation. An increase is an announcement, and announcements get answered rather than discussed.

6. Start using the eleven words this week. The next time you do something outside scope, say so as you do it. This one's on me, it's outside what we set up. That sentence keeps a favour from turning into a clause, and nobody has ever found it difficult to hear.

The agency operator with the CRM he has maintained for free for a year is not going to get that year back, and the sooner he accepts that, the better his next conversation goes. What he can do is stop the clock today, so that in eighteen months he is not writing the same message about something else he started doing in a good mood on a Tuesday.

Sources

Frequently asked

Can I ever charge for work I have already been giving away?

Sometimes, and only at a boundary with a reason of its own attached. A renewal, a scope review, a change in what the client is asking for. Billing backwards inside a live arrangement asks the client to agree they have been taking advantage of you, which is a claim about them rather than about the work, and it is the version most likely to end the relationship.

What if the client gets upset even when I only price the next request?

Some will, and that reaction is information. Upset at a clearly stated boundary usually means the boundary arrived later than it should have. Say the number before the work rather than after, keep the tone level, and accept that the first two of these conversations are the hardest ones you will have.

Isn't absorbing extra work sometimes worth it?

Often, early. A favour in the first months of a relationship buys goodwill at a moment when you have little else to offer, and I would do it again. The condition is that it stays visible: named as a favour when you do it, and given an end. An unnamed favour with no end is not generosity, it is a term of a contract you did not write.

How do I know whether something is really out of scope?

Ask whether you would have quoted it if it had appeared in the original brief. If the answer is yes, it is extra, no matter how small it feels now. Founders systematically misjudge this because the item grew gradually and each increment was genuinely trivial, which is exactly how the whole thing accumulated in the first place.

Does this apply outside professional services?

Yes, and the trades version is more expensive because the work is physical. The extra fixture, the second trip, the small repair noticed while on site. Same mechanism, same accretion, and the same correction problem eighteen months later when the client has come to expect the second trip as part of what you do.

Ready to look at the architecture honestly?

If you can list four things you do for a client every month that nobody has ever paid you for, the pricing conversation is not the one to have first. Book the conversation. We'll tell you what we see, and whether the work we do fits where you are.

Book the conversation →