Start learning free

Case study · ecommerce · revenue

The bundle that makes 'buy more' easy: Raising order value without tricks

An interactive case study: a coffee shop finds repeat buyers order one bag at a time, and tests honest volume bundles on the product page to raise average order value.

WXM 1 · Wonder (Look)WXM 2 · Architect (Plan)WXM 3 · Validate (Test)WXM 4 · Expand (Grow)
Teaching example
By Phil the CRO Guy

Phil the CRO Guy

Your CRO guide at Mentorele

Phil walks you through every lesson and case study. He asks the questions you might be too polite to ask, and he's happy to admit when a test didn't win. Every lesson he presents is checked by a real CRO expert.

Learn more about MentoreleMeet Phil

Expert-reviewed by Filip Doneski

Filip Doneski

CRO Expert and Strategist · Partner and Lead CRO Strategist at Wavesy

10+ years in conversion optimization and experimentation for DTC, retail and SaaS. Reviews every Mentorele lesson for accuracy before it’s published.

A/B testingConversion rate optimizationDigital psychology and persuasion

Filip’s expertise at WavesyLinkedIn profile

Updated

Slide 1 of 11

Hilltop Beans: loyal customers, small baskets

Hilltop Beans roasts and ships coffee. Customers love it and come back. But almost every order is a single 250 g bag, and every order pays delivery.

The team wants a higher average order value. Phil wants to know if customers want that too.

The best AOV ideas make buying more feel like the customer's idea, because it is.

Slide 2 of 11

WXM 1 · Wonder (Look)

Wonder: what the numbers show

Order data from the shop backend, checked against analytics.

  1. Most orders contain one bag.
  2. Repeat customers reorder every 3 to 4 weeks, each time paying delivery.
  3. Very few shoppers change the quantity selector from 1.
Order size distribution from the shop backend Original wireframe, not a screenshot.
  1. Quantity untouched. The quantity selector is almost never used.

Slide 3 of 11

WXM 1 · Wonder (Look)

Wonder: what customers say

The team reads reviews and sends a two-question survey to repeat buyers.

  1. Reviews: "I always run out before I remember to reorder."
  2. Survey: many repeat buyers say they'd buy two or three bags at once if it saved on delivery.
  3. Several didn't know the shop offered free delivery over a threshold.

They're not saying 'give me a discount'. They're saying 'make it easy to not run out'.

Slide 4 of 11

WXM 1 · Wonder (Look)

Wonder: what the shopper sees today

A single price and a small quantity selector. The free-delivery threshold is only mentioned in the footer.

Control (A)
Product page, current version Original wireframe, not a screenshot.
  1. Hidden option. Buying more is possible, but nothing explains why you would.

Slide 5 of 11

WXM 2 · Architect (Plan)

Architect: from evidence to a hypothesis

Because repeat buyers say they'd stock up to save on delivery but rarely change the quantity, we believe replacing the quantity selector with three clear options (1, 2 or 3 bags, with the real per-bag price and free delivery shown) will increase revenue per visitor, because it makes the bigger order easy to see, compare and choose.

Your turn: Which primary metric fits best?
Show Phil’s answer

Revenue per visitor. AOV could rise while fewer people buy at all. Revenue per visitor captures both effects in one number.

Three options, not ten. Too many choices can stall a decision when shoppers aren't sure what they want.

Slide 6 of 11

WXM 2 · Architect (Plan)

Architect: an honest anchor

The single-bag price becomes the reference point for the bundles. That's anchoring at work, and it's honest only if every number is real.

  1. Real: 3 bags are €32, which is genuinely €4 less than three single bags, plus free delivery.
  2. Not allowed: an invented 'was €45' price, or a 'most popular' badge that isn't based on sales.
  3. Anchoring is a robust effect, which is exactly why it needs care. [Source](#src-furnham-boo-2011)

Slide 7 of 11

WXM 2 · Architect (Plan)

Architect: how does it score on PXL?

A visible change above the fold on the main product pages, backed by survey and backend data. It changes how prices are displayed, so it needs some build and checking with finance.

PXL score: 9 / 14

Is the change above the fold?Yes1
Is it noticeable within 5 seconds?Yes2
Does it add or remove an element?No0
Is it designed to increase user motivation?Yes1
Does it run on high-traffic page(s)?Yes1
Does it address an issue found in user testing?No0
Does it address an issue found in qualitative feedback (surveys, polls, interviews)?Yes1
Does it address insights found in digital analytics?Yes1
Is it supported by mouse-tracking heatmaps or eye tracking?No0
Ease of implementationUp to 8 hours2

PXL by CXL. A high score ranks an idea; it does not predict a win.

Slide 8 of 11

WXM 3 · Validate (Test)

Validate: the variation

Three tappable options replace the quantity selector. The middle one is selected by default only if it really is the most bought option during a pre-test check; otherwise the single bag stays the default.

Control (A)
A: quantity selector Original wireframe, not a screenshot.
Variation (B)
B: three clear bundle options Original wireframe, not a screenshot.
  1. Easy to compare. Per-bag price and delivery are visible for each option, using real prices.

Slide 9 of 11

WXM 3 · Validate (Test)

Validate: the test plan, written before launch

Fixed in advance:

  1. Primary metric: revenue per visitor.
  2. Guardrails: conversion rate, margin per order (the discount costs money), returns.
  3. Watch later: reorder interval. Bigger orders might mean less frequent ones, so read 60-day revenue per customer after the test.
  4. QA: prices and per-bag maths correct in every currency; options work with a keyboard and screen reader.
  5. Decision rule: ship if revenue per visitor and margin improve with conversion holding.

AOV tests can fool you. Customers who stock up today may simply order less often. Plan to look again in two months.

Slide 10 of 11

WXM 4 · Expand (Grow)

Expand: whatever the result, the cycle continues

This is a teaching case, so there's no result to report. Here's what each outcome would lead to.

  1. Win: roll out to other coffees, then test a subscription option for the same customers.
  2. Inconclusive: maybe the saving is too small to notice. Research what saving would matter.
  3. Loss: check whether conversion dropped. Perhaps the options made a simple purchase feel complicated.

Slide 11 of 11

Your turn

Look at your order data: how many orders have exactly one item? If it's most of them, ask repeat customers why. The answer might be a bundle, a subscription, or just a clearer free-delivery threshold. Read anchoring before you design the prices.

What this case teaches

  • Wonder used backend order data, not just analytics, to see order size and reorder timing.
  • Architect applied anchoring and choice overload carefully: real prices, three options, an honest default.
  • Validate picked revenue per visitor and planned a longer look at reorder behaviour.
  • Expand linked the result to a bigger question about subscriptions.

Hilltop Beans is a made-up shop used for teaching.

Frequently asked questions

What metric should an average order value test use?

Revenue per visitor, which combines conversion rate and order value. AOV alone can rise while fewer people buy, and conversion rate alone can rise while orders get smaller.

Is Hilltop Beans a real company?

No. It's a made-up shop used to teach the method, so there's no real result to report.

Sources (4)

  1. A literature review of the anchoring effect. Furnham and Boo, Journal of Socio-Economics 40(1), 2011. Accessed 12 November 2026.
  2. Choice overload: A conceptual review and meta-analysis. Chernev, Böckenholt and Goodman, Journal of Consumer Psychology 25(2), 2015. Accessed 12 November 2026.
  3. WXM: the Wave Experimentation Model. Wavesy. Accessed 4 November 2026.
  4. PXL: A Better Way to Prioritize Your A/B Tests. CXL. Accessed 4 November 2026.

Learn the principles behind this case

Search lessons, guides, case studies and templates. Press Esc to close.