Your Promo Code Field Sends Buyers to Google: A Trial-to-Paid Case Study
Your Promo Code Field Sends Buyers to Google: A Trial-to-Paid Case Study
There's a particular silence on a founder call that happens right after you ask, "why do you think people aren't upgrading?" It's usually followed by a theory about price. It's almost never about the checkout form itself.
This is a case study of one field. Not a redesign, not a new pricing page. One text input in an upgrade checkout, and what happened when we stopped putting it in the buyer's face.
The setup: a nine-person team, a trial, and one text input
Context first, because generic advice is worthless without it:
- B2B scheduling tool, nine people, self-serve SaaS
- 14-day trial, no card required up front
- Upgrade path: pricing page → plan card → checkout with card details → confirm
- Two paid tiers, monthly/annual toggle
- The team ran a lot of promo codes: launch-list codes, a podcast code, two partner codes, and apology credits for a calendar sync outage
The checkout itself was fine. Card fields, an order summary, a pay button. The one thing that wasn't fine was a text input sitting directly under the order summary:
Promo code [__________] [Apply]
That field was the most-interacted-with element on the page that wasn't the pay button. Which is not a good sign. It means people were stopping to do something before paying.
We're not going to hand you a fabricated benchmark. The lift was big enough that the founder asked us to double-check the analytics were wired correctly, and the "how do I apply my code" support tickets went to roughly zero. The mechanism is the useful part, not the percentage.
Why a promo code box quietly kills trial-to-paid
Three separate things happen when a buyer sees that field, and none of them are about money.
1. It sends them out of your checkout. They open a new tab, search the product name plus "discount code," and land on a coupon aggregator, a Reddit thread, or a competitor's comparison page. You just asked a person who was ready to pay to go read about your competitors. That's friction in its purest form: a step that has nothing to do with the transaction you're trying to complete.
2. It's a trust tax. A visible discount field tells the buyer, implicitly, that the price on this page is negotiable and they might be the only one paying full price. That's a clarity problem — the price stops being the price and becomes an opening offer. Nielsen's visibility-of-system-status heuristic applies here in an uncomfortable way: the system is telling them the status is "maybe you're getting ripped off."
3. It manufactures error dead-ends. Codes expire. Codes are case-sensitive. Codes were scoped to annual only. Codes were for a tier they didn't pick. Every one of those produces a rejection message at the worst moment in the funnel — the second before someone types a card number. Feedback that arrives as a no at the point of payment is worse than no feedback at all.
There was a fourth, internal cost: support was manually applying codes in the billing system, which meant the checkout and the invoice occasionally disagreed. Nothing erodes trust faster than a receipt that doesn't match the page you paid on.
The one change: auto-apply, collapse, show the math
The fix wasn't "delete promo codes." People had codes. Some had been promised a discount in writing. Removing the field outright would have created a different, louder problem.
So we shipped three moves as one change:
- Auto-apply on load. If the account's email or workspace is tied to an active code — launch list, partner referral, apology credit — apply it before the page renders. No input, no Apply button, no confirmation toast.
- Collapse the manual field. A small text link under the order summary: Have a code? It reveals the same input. It's still there. It's just not the first thing a ready-to-pay buyer sees.
- Show the discount as a line item in the math, not as a floating success message. A discount that appears as a toast disappears. A discount that appears as a line in the receipt becomes part of what the buyer is reading and agreeing to.
Point 3 is the one teams skip. If the discount isn't in the summary, the buyer does the subtraction themselves and gets it wrong, or assumes it didn't apply.
Before and after: the actual copy rewrite
Before
Order summary
Pro plan — $75.00 / month
Total due today: $75.00
Promo code [__________] [Apply]
[ Start subscription ]
Three problems: the promo field sits between the price and the button, the button doesn't say what happens, and "total due today" never tells you what happens next month.
After
Order summary
Pro plan (monthly) $75.00
Launch code LAUNCH25 −$18.75
────────────────────────────────────────────
Due today $56.25
Then $75.00/month starting 14 Mar
Have a code?
[ Start subscription — $56.25 today ]
What changed, and why each part matters:
- The discount is a line item, not a surprise or a mystery
- The renewal price and date are stated before the button, not in an email a month later
- The button repeats the amount, so the last thing they read is the number they're agreeing to
- "Have a code?" is a link, not a field — it doesn't compete with the CTA for attention
- Plan naming matches the /pricing page exactly, so nobody has to wonder if "Pro" on checkout is the same "Pro" they clicked
Hierarchy check: before, there were two clickable things and neither was obviously the main one. After, there's one primary action and one secondary escape hatch. That's the whole point.
A P0/P1/P2 playbook for your own upgrade checkout
Run this against the live page in production, not the Figma file.
P0 — this week
- Any input sitting between the order summary and the pay button that isn't payment info: move it below, collapse it, or delete it.
- Auto-apply every discount you can resolve from the account. If a human in support has to apply it manually, it should be automatic.
- Put the next charge amount and date above the pay button.
- Make the pay button state the amount.
P1 — next two weeks
- Kill error dead-ends. If a code fails, say why in one line and leave the input filled so they can fix it. Never clear the field.
- Log every failed code attempt with the raw string. You'll find expired partner codes and case-sensitivity bugs within a day.
- Test the back button: checkout → pricing page → forward again. Does the plan, the interval, and the applied code survive?
- Make the confirmation email match the checkout line items exactly.
P2 — quiet afternoon
- Compare redemption rate before and after collapsing the field. If it drops sharply, your codes were being discovered by accident, not by intent — that's a distribution problem, not a checkout problem.
- Apply the same lens to your annual toggle: does switching intervals preserve the discount, or silently reset it?
- Add a short "what you're paying for" line to the summary. One sentence beats a feature list at this stage.
If you'd rather have that list handed to you than assemble it from scratch, run a free audit on your upgrade checkout — it returns the same P0/P1/P2 structure in a few minutes.
How to measure it without lying to yourself
Two traps, and both are common on small teams.
Trap one: page-level conversion instead of flow-level. Someone abandons checkout, comes back twenty minutes later on their phone, and converts. Page analytics count two sessions and one conversion; the page looks worse than the flow is. Measure trial-to-paid by trial-start cohort, not by checkout session.
Trap two: attributing a seasonal bump to your change. If you shipped in the same week as a launch, a newsletter, or a price change, you don't know anything yet. Hold a control group if you have the traffic, or at minimum compare equal-length windows and check whether traffic mix shifted.
The metric that actually matters: share of trials reaching paid, by cohort, plus a secondary on support tickets tagged billing/code. If the code tickets go to zero and cohort conversion doesn't move, you fixed a support problem, not a conversion problem — and you should say so out loud instead of claiming a win.
If you want the pricing-page side of this too, there are more teardowns on the /blog.
You don't need a redesign to move trial-to-paid. You need to find the one element in the flow that's asking the buyer to do something other than buy, and get it out of the way. Start a free FlowAudit at /signup and you'll get a prioritized P0/P1/P2 fix list for your own checkout in minutes — no call, no consultant, no forty-page deck.