The Birth of the Price Tag: a Trust Machine

Price Tag

Most mornings, around sunrise, the dogs and I head down Sixth Avenue. The route runs past a row of enormous cast-iron buildings that now house big-box stores and gyms. In the late 1800s, they were the dry goods palaces of Ladies’ Mile, the most important shopping district in America.

The dogs care more about the hydrants and scrounging for “street food”. I’ve become fascinated by something else these buildings represent: the moment American retail stopped haggling. 

And I think B2B software is about to relearn the lesson these merchants figured out 150 years ago.

Why haggling died

In 1858, Rowland Hussey Macy opened a dry goods store at Sixth Avenue and 14th Street, the southern gateway to Ladies’ Mile. He borrowed a Quaker principle: every customer pays the same price for the same item. He posted those prices, advertised specific prices in newspapers, took cash only, and backed it with a money-back guarantee.

That was radical. Haggling was the norm, and every sale was a negotiation that depended on who was buying and who was selling. Merchants didn’t abandon it out of generosity. They abandoned it because it broke under three pressures:

  1. A new buyer. A growing urban middle class, many of them women shopping on their own, who had no interest in out-negotiating a clerk they didn’t know.
  2. Scale. Stores with hundreds of clerks could not trust each one to negotiate margin on every sale.
  3. Trust. A posted price was a promise: everyone pays the same, and you can see it before you commit. The guarantee went further: if it doesn’t deliver, bring it back.

The price tag wasn’t a pricing model. It was a trust machine. And it let these stores scale into the department stores we know today.

We’ve reinvented the haggle

For two decades, B2B software had its own price tag: the seat. It was simple, posted, and easy to budget. AI agents are breaking it, because when software does the work, the number of humans logging in stops tracking the value delivered.

The market is moving fast. Kyle Poyar’s 2025 survey of 240+ software companies found seat-based pricing fell from 21% to 15% in a single year, while hybrid models jumped from 27% to 41% (Growth Unhinged). Bain’s analysis of 30+ SaaS vendors adding generative AI found about 65% layered an AI usage meter on top of seats, and none had fully moved to usage or outcome pricing (Bain, 2025).

Here’s the problem nobody wants to say out loud. Seats, then tokens and credits, then outcomes: each step is more aligned in theory and more opaque in practice.

  • Usage and tokens are hard to forecast. Gartner has warned that AI cost estimates can be off by 500% to 1,000% (The Register, 2024). IDC predicts large enterprises will underestimate AI infrastructure costs by 30% (CIO.com, 2025).
  • Outcomes sound perfect, until buyer and seller have to agree on what counts as an outcome, how it’s measured, and who gets credit.
  • Hybrid often means a seat price, plus a meter, plus a negotiated commit, plus a true-up.

We’ve gone back to the bazaar. Every deal is a custom negotiation, and the buyer can’t see the price until they’re deep into the sale. 

These are the same three pressures as 1858: a new buyer (the CFO and procurement scrutinizing every renewal), new scale (agents doing work at volumes no seat count captures), and a trust deficit.

What the merchant princes knew

The debate over seats versus tokens versus outcomes misses the point. Macy’s didn’t win because it picked the right unit of measure. It won because buyers could predict what they’d pay and verify they got what they paid for.

That’s the real test for any pricing model today:

  • Can the buyer predict it? If a CFO can’t forecast next year’s bill within a reasonable range, the model will be fought, capped, or quietly replaced at renewal.
  • Can the buyer verify it? If value isn’t measured and reported, outcome pricing becomes a dispute waiting to happen.
  • Does the vendor share the risk? The money-back guarantee was the original risk reversal. Outcome commitments are the modern version, but only when they’re backed by transparent measurement.

The winning model isn’t per seat, per token, or per outcome. It’s whichever one the buyer can trust.

The modern price tag is a value model

If you’re changing how you price, the pricing page is the easy part. The hard part is rebuilding the trust the seat used to provide. Four moves for value-led vendors:

  1. Show the math before the sale. Give buyers a transparent ROI and TCO model that connects your pricing unit to their outcomes, with every assumption visible and editable. That’s today’s posted price.
  2. Make cost predictable. Pair usage or outcome meters with ranges, caps, or committed tiers so the CFO can budget with confidence.
  3. Agree on the outcome up front. Define what counts, how it’s measured, and the baseline, before the contract, not at the true-up.
  4. Prove it after the sale. Report realized value on a regular cadence. Customer success becomes the guarantee, and value realization becomes the renewal case.

Vendors who do this won’t just justify a new pricing model. They’ll differentiate on trust while competitors haggle.

Back on the street

By the time the dogs and I loop back to 23rd Street, the sun is up on those cast-iron facades. 

The merchants who built them are long gone, but the lesson is still standing: buyers don’t reward the cleverest pricing model. They reward the one they can see, predict, and trust.

AI is rewriting how software is priced. The vendors who win will be the ones who bring back the price tag, in the form of a transparent value model.

Need help building the value models that make your new pricing trustworthy? 

Let’s discuss what this looks like for you:  book a one-on-one conversation.

Sources

 

Sign-up for our Newsletter