Pricing & Offer Tests for Real People
Practical experiments for pricing and offers customers actually understand—guardrails, test design, examples, and how to read the results without wrecking margin.
Dom O'Brien
11/4/20252 min read


Pricing isn’t theoretical—it’s how people perceive value in a moment. These tests keep ops sane, protect margin, and still give you signal.
Guardrails (pin these above your desk)
One change at a time. No multiverse of promos.
Hold for a full buying cycle. Don’t call winners in 48 hours.
Protect margin. Set a contribution floor and honor it.
Communicate clearly. One offer per page/email; stacking rules in plain English.
Segment cleanly. Don’t cross‑contaminate traffic sources.
Five experiments to run next
Threshold vs sitewide
Question: Do buyers spend more with “$20 off $100+” than “20% off sitewide”?
Design: 50/50 split; identical pages except for framing.
Watch: AOV, conversion rate, returns rate.Bundle vs a‑la‑carte
Question: Does a starter kit beat single items?
Design: Build a bundle with a small saving and a super clear “what’s in the box.”
Watch: AOV, attach rate, support tickets.Free shipping threshold
Question: What number drives the most net revenue?
Design: Test two thresholds for a week each; show a progress bar in cart.
Watch: AOV, gross margin, cart abandon.Price presentation
Question: Percent vs dollar saving - what’s clearer?
Design: Hold final price constant; change only how you show the saving.
Watch: Click‑through to checkout, code usage, conversion rate.Urgency placement
Question: Does a shipping‑cutoff timer near price beat a generic header countdown?
Design: A/B position only; keep style identical.
Watch: Conversion rate by device; bounce on mobile.
Read the results
Use a one‑sheet: traffic source, variant A/B, metrics (CR, AOV, RPS, MER), sample size/date range, and trade‑offs (e.g., higher AOV but more returns). Decide the winner (and take it out of test mode), log the learnings, queue the next test.
Pitfalls
Mixing promotions and then “learning” nothing.
Pushing discounts where bundles would protect margin better.
Short tests that catch a weekend quirk, not reality.
Over‑rotating on conversion rate while net revenue falls.
Pricing wins when it’s clear for customers and healthy for your margin. Start with one simple experiment this week and hold it for a full buying cycle. Read the result on CR, AOV, RPS and MER, then ship the winner, log the learning, and queue the next test.
Keep your guardrails visible and brief finance on the contribution floor and rollback plan; you’ll build a pricing playbook that compounds rather than whiplashes the business.
FAQ
Do I need an offer calendar? Yes - so ops and support aren’t surprised.
How do I brief finance? Share contribution floors and a rollback plan.
What if a test tanks? End it early; that’s a win and a cheap lesson.
Contact Us
© 2025. All rights reserved.
