I trained as an economist before I ever wrote a line of title software, and the habit that never left me is looking at the shape of a price instead of the level of it. Almost everyone negotiates the level. The shape is where the money actually is. Per-order pricing has a very specific shape: a straight line through the origin. Order number one and order number nine hundred cost exactly the same, forever. On the day you sign, that line looks like the fairest thing in the room. Give it three years and one rate move and see how it feels.

The Straight Line Is Seductive for a Reason

Per-order pricing wins deals because it’s legible. You can compute your bill exactly, every month, with arithmetic you can do in your head. There’s no true-up, no shelfware, no seat count to defend after a resignation. Close nothing, pay nothing. For an owner who once signed a seat-based contract at a headcount they no longer have, that’s worth real money. It also matches cost to activity — your software becomes a variable cost that breathes with revenue. Textbook. All of that holds inside a single month. But you buy a production system across a cycle, and title cycles are violent.

What That Line Does Across the Cycle

Here’s the asymmetry. A fully variable cost protects you on the way down — that’s genuine, and in a bad year it’s the difference between trimming and bleeding. But the same property that shields you in the trough takes a fixed percentage off the top of every good month you will ever have. You have sold the vendor a permanent share of your upside in exchange for insurance on your downside. That can be a fine trade. It’s worth knowing you made it. From the other side of the table, the vendor’s cost to serve your nine hundredth order is close to nothing — same as their first. Your price didn’t move. The gap between what that order costs them and what it costs you is pure margin, widest precisely when you’re busiest. Your best month is, by construction, their best month. That’s not a conspiracy. It’s just what a straight line does.

That gap isn’t incidental to how software works; it is how software works: high fixed cost, near-zero marginal cost. The build is expensive; the hundred-thousandth transaction through an already-built pipeline is not. A price that doesn’t bend as volume rises ignores the economics of what’s being sold. Scale efficiency exists in this business; the only question is who keeps it. That’s what tiers are for — not goodwill, but the mechanism that pushes some of that efficiency back to the customer instead of the vendor keeping all of it. A vendor who won’t tier at any volume, ever, is telling you which of those two they intend. Believe them.

Don’t Compare Vendors at Your Average Month

This is the part most shops get wrong. Pull your monthly order counts for the last two or three years and do not average them. Look at the distribution — worst month, median month, best month, and how often you land near the top of that range — then price each structure at each of those points, not at the mean. For a straight-line vendor, your bill at your average month is your average bill; the math is exact. For a tiered vendor it isn’t — because the price bends, your true average bill comes in below what the average-month calculation shows. Evaluating both deals at your median is therefore not neutral — it is systematically biased in favor of the linear vendor, and the more volatile your volume, the bigger the bias. Title volume is about as volatile as it gets.

Two deals that look identical at your median can diverge badly at your peak, and your peaks are where your margin actually gets made. The number you want is the break-even: the volume at which the tiered structure’s step-down overtakes the flat rate. Find it, then ask how many months out of the last thirty-six you spent above it. If the answer is “a lot,” you’ve been paying for a structure you outgrew.

Where Per-Order Is Still the Right Answer

Let me be straight here, because the argument doesn’t work if I overreach. If you’re a small shop, or your volume swings unpredictably, or you’re seasonal enough that a quiet stretch is normal, per-order is probably correct for you. Same if you want zero commitment on principle — that’s a legitimate preference, not a failure of sophistication. And if you never come near the first tier break, tiers are decoration. The claim isn’t that per-order is a scam. It’s that per-order stops making sense above a volume you may have already passed without noticing, because nobody re-examines the pricing structure in a busy year.

The Trap on the Other Side

Tiers have their own failure mode, and it’s real. The moment a tier comes attached to a minimum or an annual commitment, the vendor has handed the risk back to you. A guaranteed volume in a down year is a fixed cost you cannot shed — exactly the thing per-order was protecting you from. You didn’t get a better structure; you traded one exposure for another and paid for the privilege. So price the minimum as what it is — a cost, not a footnote. If a tier requires a commitment, take your worst observed month, work out what you’d be paying for orders you never ran, and add that to the quote. Sometimes it still wins. Often it doesn’t. The structure you actually want is the step-down without the floor.

Where the Software Earns Its Keep

This is the part I built a company around, so weigh it accordingly. Autopilot is priced in volume tiers rather than on a flat per-order line, for the cost structure I described above. The work it does — automated, high-repetition production work that scales with your order count — genuinely gets cheaper for us to deliver as your volume rises. Pricing it as though it didn’t would mean charging you for an efficiency you’re the one creating. So the rate steps down as you grow: your peak months cost proportionally less per order than your quiet ones, because that’s what’s actually happening underneath.

Every pricing conversation gets framed as a negotiation over rate. It almost never is. It’s a negotiation over who captures the value of your best year — and that question gets settled by the structure long before anybody argues about the number.

Run your real volume against Autopilot’s tiers →