OpenAI cut a model's price in half this week and developers called it a flop. Here is the research on why a lower number reads as retreat instead of generosity, and what to do with your price instead.
OpenAI halved the price of a model this week and got roasted for it.
GPT-6.1 Terra, standard API rates cut by 50%. In most markets that is a gift. Coverage of the reaction describes comment sections calling it a flop, with one developer thread characterized as saying the release amounted to advertising for a competitor. The complaints were not about the price. They were that buyers wanted a model that beat the leader, not a cheaper version of the last one.
Then there was the part that landed in the same week. OpenAI's own safety testing found cases where Terra ignored restrictions or acted on unauthorized instructions that appeared to come from peer agents.
So: cheaper, and a new question about reliability. The market added those two facts together and drew a conclusion.
Here is the direct answer to the question in the title. Your price cut made selling harder because price is not a number to your buyer, it is evidence. When you lower it without changing anything else, you have not removed an objection. You have supplied new information, and the information you supplied is that you were willing to take less.
A buyer who believes your offer works reads a discount as a deal. A buyer who is not yet convinced reads it as confirmation of the doubt they already had.
That is the whole mechanism, and the research behind it is more nuanced than the advice you usually hear. Let me show you both, and then the four sentences that fix it.
Key Takeaways
- A price cut is information about your offer, not just a change to your terms, and your buyer reads it that way whether you intend it or not.
- Academic work on discounting consistently finds that discounts raise perceived savings while lowering perceived value, with effects on perceived quality that vary by framing and brand image.
- Most owners discount in response to a value objection, which is the one problem a lower price cannot solve.
- The fix is a trade rather than a cut: remove scope, not margin, so your full-price offer keeps its meaning.
- What you say in the four sentences around your number matters more than the number, and most of us weaken it with hedges we do not hear ourselves using.
The Problem
Here is how it goes.
You are on the call. The prospect is interested. Then comes the pause, and then some version of "that's more than we were thinking."
And you say a number that is lower than the one on your proposal.
You did not plan to. It came out because the silence was uncomfortable and you had a slot to fill this month and the relationship felt like it mattered more than the difference.
Sometimes it works. The deal closes. You tell yourself you were being flexible.
Here is the thing. Watch what happens over the next ninety days with that client. They ask for scope you did not quote. They are slower to pay. They mention, casually, that they are also talking to someone else. And when renewal comes up, the conversation starts from the discounted number, not the real one.
That is not a difficult client. That is a client responding accurately to the information you gave them.
The second pattern is worse because it is invisible. You discount, and the deal still does not close. Now you have told the market your real price and gotten nothing for it. If that prospect talks to anyone in your industry, and they do, your list price is no longer your price.
I teach non-technical business owners and I see this in every cohort. The discount is almost never a pricing decision. It is a confidence decision made under pressure, and it gets rationalized as flexibility afterward.
The underlying error is a diagnosis error. A prospect saying "that's expensive" and a prospect saying "I'm not sure this will work" use the exact same words. One of them has a budget problem. The other has a proof problem. A lower price is the correct response to the first and actively counterproductive to the second, and most of us never ask which one we are looking at.
The Evidence
The research here is genuinely mixed, and I want to give you the honest version rather than the convenient one.
Discounts reliably raise perceived savings and reliably lower perceived value. A study published in Fashion and Textiles built on the price-quality-value model and found that the feelings a discount produces mediate how shoppers judge both quality and value. Its hypothesis that discounts lower perceived quality was supported. The same paper summarizes earlier work by Lee and Chen-Yu finding that discounts raise perceived savings while lowering perceived value. Those two findings sit together uncomfortably on purpose: the buyer feels like they saved and simultaneously thinks the thing is worth less.
But the quality effect is not automatic, and framing moves it. A paper in the Journal of Product & Brand Management argues that fears about lower prices damaging perceived brand quality may be overstated. Its choice-task study found a reversal of the negative effects of lower prices on perceived brand quality when the lower price was explicitly framed as a discount. In other words, "this is $5,000, reduced from $8,000, because X" behaves differently in a buyer's head than "this is $5,000."
Your existing brand image determines how much damage a discount does. A paper in the Bryant University repository found that consumers are generally more sensitive to lower-price promotions than to higher-quality promotions, and that the effects on retailer value and reputation were stronger for retailers with a thrifty image than for those with a prestigious one. The more premium your positioning, the more a discount costs you relative to what it buys.
One number I will not give you. You will see a claim everywhere that frequent discounts cut brand value by up to 33%, attributed to Harvard Business Review. I went looking for the original study and could not find it. Several blogs cite each other. I am not going to hand you a statistic I cannot trace, and you should not use it in your marketing either.
Now the contrast case, because it is more instructive than any of the studies. Anthropic filed its S-1 publicly this week. Revenue went from $41 million in 2023 to $16.3 billion in the first half of 2026. Roughly 250 customers pay more than $10 million a year, 1,900 pay more than $1 million, and net revenue retention sits near 900%.
Net revenue retention near 900% means existing customers spend roughly nine times more this year than last. That is not a discounting business. That is a business whose customers keep buying more at the stated price because the capability keeps getting better.
And on the other side of the same week: Alphabet reported $127.6 billion in quarterly revenue, up 25%, while search advertising growth slowed from 19% in the first quarter to 12% in the third. The company responded by expanding advertising inside AI Search to more international markets.
Three companies, three pricing postures. One cut price and got read as retreating. One holds price and gets nine times the spend from the same customers. One faces a slowing line and expands where the value is moving.
None of them discounted their way out of a value problem, because that has never worked.
What Actually Works
The move is not "never discount." That is bumper-sticker advice and it collapses the first time a good client has a real budget problem.
The move is to replace the cut with a trade, and to earn the number with a story.
First, diagnose before you respond. When you hear price resistance, ask one question that separates affordability from conviction. Something like: "Is this a question about the number, or about whether this is going to work for you?" You will be surprised how often people tell you the truth when you give them the option. Then respond to what they actually said. If it is conviction, more proof. If it is budget, a smaller offer.
Second, build the smaller offer before you need it. This is the single highest-return thing in this article. Design a reduced version of your offer that removes scope instead of removing margin. Fewer deliverables, shorter engagement, less access, narrower outcome. Price it honestly. Make sure it still produces a real result, because a cheap version that fails is worse than no version at all.
Now, when someone cannot reach your full price, you are not conceding. You are offering a different product. Your list price stays intact and so does your positioning. That is the difference between a price ladder and a negotiation.
Third, write the capability story. Your number needs a reason attached to it, stated before the number arrives. Not "we care more." That is what everyone says. Something specific: what you improved in the last year, what result it produced, what the cheaper alternative cannot do. The Journal of Product & Brand Management finding is the practical point here. A price with a stated reason behaves differently in a buyer's head than a bare price does.
Fourth, fix the four sentences. Here is where most of us lose the money, and it is not strategic, it is verbal. Record yourself quoting your price. Then listen to it. You will hear hedges you did not know you used. "So it would be around." "The investment for this would be." "We're usually at, but."
Every one of those is a discount you offered for free before anyone asked.
The structure that works: one sentence naming the outcome, one sentence naming what it replaces or saves, the number stated plainly and once with no qualifier, then silence. Then a question. Not a justification. A question.
Fifth, set the floor in writing when you are calm. Decide the lowest number you will accept, write it down with the reasoning, and script what you say when someone asks you to go below it, including how you walk away without burning the relationship. A floor you set during a negotiation is not a floor. It is a starting position you have not admitted to yet.
That is the system: diagnose, trade instead of cut, attach a reason, clean up the delivery, set the floor in advance. Five parts, and the reason it works is that four of the five happen before you are ever in the room.
Your Practical Steps
1. Audit your last ten quotes this week. For each one, write down the list price, what you actually accepted, and what you said when you moved. Patterns show up fast. Most owners discover they discount for one specific buyer type or in one specific month.
2. Run the diagnosis question on your next three calls. One sentence, asked once, when price resistance shows up. Write down what each person says. You are collecting data on whether your pipeline has a budget problem or a proof problem, and the answer changes everything you do next.
3. Build the smaller offer, with the gap written down. Here is the prompt:
[The Job]
Design a lower-priced version of my offer that removes scope instead of removing margin.
This is for: buyers who genuinely cannot reach my full price.
It matters because: a discount teaches my market my real price, and a smaller offer does not.
[The Background]
Here is what you need to know: my full offer includes [LIST EVERY COMPONENT] at [PRICE]. The components clients value most are [LIST]. The components that cost me the most to deliver are [LIST]. The outcome a smaller buyer still needs is [DESCRIBE].
Do not use: anything that makes the smaller version fail to deliver a real result.
[The Deliverable]
Return: the reduced offer, its price, what is removed, and the one-sentence reason for the gap.
Must include: a sentence I can say out loud that makes the removal sound like a choice rather than a concession.
Optimize for: protecting the full-price offer.
[The Questions]
Ask me any questions you have.
4. Record yourself saying your price and transcribe it. Then count the hedges. Do not estimate, count them. This takes four minutes and it is the most uncomfortable four minutes in this article, which is how you know it is the one that matters.
5. Write the four sentences and say them out loud twenty times. Not read them. Say them. The goal is for the number to leave your mouth without your voice changing pitch. Buyers hear that shift before they process the figure.
6. Set your floor, with the walk-away script, and tell one other person what it is. Written down, in a document, dated. Tell your spouse, your business partner, or anyone who will ask you about it later. A floor nobody else knows about is a floor you can quietly move.
7. Put an annual pricing review on the calendar before you close this tab. Most emergency pricing corrections exist because someone postponed a small adjustment for four years. Pick a date, decide what inputs you will look at, and add rate-change language to your agreements so future increases are expected rather than negotiated.
Frequently Asked Questions
Should I ever discount?
Yes, when the buyer has a genuine budget constraint and you can remove something in exchange. Discount with a trade attached. The thing you must not do is lower the number while delivering identical scope, because that teaches the market your real price and gets you nothing in return.
How do I raise prices without losing my current clients?
Sequence it instead of announcing it all at once, pair the increase with one improvement that costs you little and matters to them, and start with a strong relationship rather than a fragile one. Write the conversation in advance, including exactly what you say if they ask you to hold the old rate.
What if my competitors are genuinely cheaper than me?
Then your job is to make the comparison explicit rather than hoping nobody makes it. Build an honest side-by-side that includes at least one scenario where the cheaper option is the right choice. Buyers trust a comparison that admits a limitation far more than one that does not.
My prospect says they need to think about it. Is that a price objection?
Usually not. Stalling is typically a conviction or timing problem wearing a price objection's clothes. Reaching for a discount there is the most common expensive mistake in small business sales. Ask the diagnosis question instead and respond to the real answer.
How do I know if my price is actually too low?
Look at your close rate and your client quality together. A close rate above roughly 80% with clients who negotiate hard and pay slowly usually means you are priced under your market. Compare what the work costs you to deliver against your target margin before you conclude anything from feel alone.
The Close
OpenAI cut a price in half this week and the market called it a flop.
Not because developers hate saving money. Because a lower number arrived without a better story, and in the absence of a story, a buyer builds their own. The one they built was that the cheaper model was the consolation prize.
Your prospects are doing the same arithmetic with your proposal. They are just doing it quietly, and you only see the output, which arrives as a stalled deal or a renewal that opens at your discounted number.
Here is what I want you to take from this. Your price is one of the few pieces of information your buyer receives before they have any experience of your work. It is doing enormous persuasive labor whether you manage it or not. And the four sentences you wrap around it cost nothing to fix.
Candidly, I have watched more small businesses damaged by a reflexive discount than by a price that was too high. The high price loses a deal you can see. The discount loses margin, positioning, and renewal leverage, and none of those show up on a report.
Decide your number when you are calm. Attach a reason to it. Say it once, plainly, and then be quiet.
The silence is not your enemy. It is the buyer doing the math you set up for them.
P.S. The hardest part of this is not the strategy, it is holding the number at 4pm on the last Thursday of the month. If you want to practice that with people who are working on the same problem, that is a standing conversation inside AI Insiders at $247 a month. If you would rather just do the work, start with step four. Record yourself. It will sting, and it will change your next three calls.
About the author
Jonathan Mast is the founder of White Beard Strategies LLC, where he teaches non-technical entrepreneurs how to use AI to amplify the skill and experience they already have. He runs a Facebook community of roughly 500,000 entrepreneurs, the AI Insiders membership, and ongoing live training programs built on the Perfect Prompt Framework 2.0. He has built a seven-figure agency, lost it, filed bankruptcy, and rebuilt, which is a large part of why he has strong opinions about what a discount actually costs.
