How to run a pricing decoy test that increases average deal size by 12% using three controlled offers

I ran a small pricing experiment last quarter that lifted average deal size by ~12% within four weeks — using a simple decoy setup with three controlled offers. I want to walk you through the exact approach I used, why it works, how to design the test, what metrics to track, and common pitfalls to avoid. This is the playbook I use when I need fast, measurable wins in B2B and high-consideration B2C offerings.

Why a three-offer decoy test?

Decoy pricing leverages psychology: customers compare options relative to each other, not just against an absolute value. Presenting a middle and high option alongside a clearly inferior “decoy” can nudge buyers toward the more profitable choice. I prefer a controlled three-offer setup because it’s:

Simple to A/B test in funnels (two variants: control vs. decoy).

Easy to explain to stakeholders and product teams.

Fast to implement in pricing pages, proposals, or checkout flows.

In practice, the decoy isn’t a deception — it’s an intentionally designed option that makes the target product look comparatively better on price-to-feature ratio.

How I frame the hypothesis

Every experiment starts with a clear hypothesis. Mine was:

Introducing a decoy option will increase the share of customers selecting the target (higher-margin) offer and increase average deal size by at least 10%.

I defined the target offer as the middle-priced plan (not the cheapest, not the most premium). That’s where most customers lean when presented with meaningful differences.

Designing the three offers

The three offers need clear positioning and measurable differences. Here’s the structure I used for a SaaS product example:

Plan	Price (monthly)	Key features	Role in experiment
Basic	$49	Core features, limited seats	Low-priced baseline
Pro (Target)	$99	Unlimited seats, advanced reporting	Target option we want to increase selection
Pro+ (Decoy)	$129	Same as Pro + minor extra (e.g., 1 extra integration)	Decoy: priced close to Pro but offers minimal extra value

Key design principles I follow:

The decoy should be priced above the target but offer only marginally more value, so the target looks like the best value.

The basic plan should be plainly cheaper and visibly less capable.

Feature descriptions must be crisp and comparable — vague language kills the effect.

Where to run the test

I run decoy tests in one of three places depending on volume and sales cycle:

Public pricing page — best for high-traffic SaaS sites and e-commerce where purchase friction is low.

Proposal templates — ideal for B2B deals with a rep involved; keeps the test within controlled outreach (same salesperson, same messaging).

Checkout flow or modal — effective for transactional products where customer chooses quickly.

When testing on pricing pages I prefer server-side experiments (e.g., via LaunchDarkly or Optimizely) so the test is consistent across pages and devices. For proposals, I randomize which template the salesperson uses and track outcomes in the CRM.

Implementation steps (play-by-play)

Set baseline: record current conversion rate, average deal size (ADS), and plan mix for at least 2–4 weeks.

Define variants: Control = existing pricing (usually two offers or three evenly spaced). Test = add decoy or replace current high-tier with decoy.

Randomize traffic: split traffic 50/50 or 60/40 depending on risk tolerance. Keep segmentation consistent (same traffic sources in each group).

Run the test for a minimum sample: aim for statistical significance but also practical time — typically 3–6 weeks for mid-volume sites, longer for low-volume B2B.

Track metrics in the analytics stack and CRM (see KPIs below).

Analyze results with confidence intervals — don’t stop early on a fluke.

KPIs and how I measure success

Primary KPIs:

Average deal size (ADS) — revenue / number of transactions.

Plan mix — percentage selecting each plan.

Conversion rate — visits → signups or proposals → closed deals.

Secondary KPIs (to watch for adverse effects):

Churn/retention at 30/90 days.

Sales cycle length (for B2B).

Support/contact rates — did more people ask questions because pricing became confusing?

Example: baseline ADS = $110, plan mix: Basic 40%, Pro 50%, Premium 10%. After introducing the decoy, Pro selection rose to 58%, Premium fell to 6%, Basic stayed ~36% and ADS moved to $123 (+12%).

Statistical considerations

Don’t rely on raw percentage changes without testing for significance. I use a two-proportion z-test for plan-share changes and t-tests for ADS. If you don’t have statistical tooling, online calculators from Optimizely or Evan Miller help.

Minimum sample guidance:

For high-traffic pricing pages — aim for at least 1,000 conversions per variant.

For proposal-based experiments — aim for at least 100–200 deals per variant, recognizing longer timelines.

Real experiment notes and copy examples

In one test I used the following microcopy changes that reinforced the decoy effect:

Pro: “Most popular — Best value for growing teams.”

Pro+: “Includes X integration — for teams that need a single extra connector.” (small visual badge, muted color)

Basic: “Essential features to get started.”

Buttons:

Pro: “Start Pro — Best value”

Pro+: “Start Pro+ — Only $30 more”

Small tweaks like labeling the middle plan “Most popular” and showing monthly/yearly toggles with savings highlighted make a measurable difference.

Common pitfalls to avoid

Creating a decoy that’s obviously inferior or ridiculous — you’ll lose credibility.

Making feature comparisons fuzzy — customers must see a clear trade-off.

Changing multiple variables at once (price + major feature set + layout) — then you won’t know what moved the needle.

Ignoring downstream metrics — a higher ADS is worthless if churn spikes.

When not to use a decoy

Decoys work best when customers compare options consciously. Avoid this tactic when:

Your audience buys on impulse and isn’t comparing features (pure commodity goods).

Your offering requires deep customization — then pricing needs to be consultative.

Quick troubleshooting

If the decoy doesn’t work:

Check that the decoy’s price gap is sensible (10–40% above the target usually works).

Confirm feature clarity — ask 5 customers if they understand differences.

Look for skewed traffic — if a new marketing channel brings bargain hunters, your plan mix will shift unpredictably.

Template: hypothesis & experiment brief

Field	Example
Objective	Increase average deal size by >10% in 6 weeks
Hypothesis	Adding a decoy option at $129 will shift customers to Pro at $99
Primary KPI	Average deal size
Secondary KPIs	Plan mix, conversion rate, 30-day churn
Traffic split	50/50 control vs. test
Duration	6 weeks

If you want, I can draft a tailored brief for your product — give me the current prices, traffic volume and where you’d place the test (pricing page, proposal, checkout) and I’ll map the concrete next steps and trigger texts for your team.