Netflix Style Pricing Pages

Netflix A/B tested at least three pricing plan layouts between 2019 and 2026. Now, in 2026, a third version is live. Run the experiment and see whether Netflix really picked the winning treatment.

MDE range 1-3%

Lessons 8

Taught
by

Ishan
Goel

Jakub
Linowski

Your existing pricing page

Choose your plan

Downgrade or upgrade at any time.

Monthly price

$9.99

$13.99

$16.99

HD available

Ultra HD available

Screens you can watch at the same time

Watch on your laptop, TV, phone and tablet

Unlimited movies and TV shows

Cancel anytime

HD and Ultra HD availability subject to your Internet service and device capabilities. Not all content available in HD or Ultra HD.

Obvious plan tabs
Scannable comparison rows
Strong selected column

Choose your plan

Downgrade or upgrade at any time.

Stacked plan cards
Benefits near price
Familiar radio state

Choose your plan

Downgrade or upgrade at any time.

Monthly price $16.99

Resolution Ultra HD

Screens 4

Laptop, TV, phone & tablet Yes

Unlimited movies & TV shows Yes

Cancel anytime Yes

Ads No ads

Compact first-screen cards
Details after selection
Light choice hierarchy

Simulated Test Run

Variant	Visitors	Purchases	Purchase rate	True effect	Measured effect	p-value

Experiment Design

Choose the structure of your experiment.

Single-Variable

Single change A/B test.

Exploratory

Several variations tested at once.

Exploratory + Confirmatory

Lower power first, higher power second.

Leap Compound

Multiple simultaneous changes in one variant.

Intensity

One change at varying strengths.

Cascade

Each change builds on the last.

Design File

Link to detailed designs or mockups with more of the specifics.

Prior Evidence Checks

To increase your odds of success, maximize impact and decrease sample size.

Should a default plan be pre-selected?

Recommended: Highest Plan Selected, worth +10.2% in Test 525. The winning variant defaulted to the most expensive option, and the lift came from anchoring rather than from the discount itself - so keep the price visible while the plan is pre-selected.

How many plans should be offered?

Recommended: More plans, worth +9.3% in Test 319. Cutting the lineup down to fewer plans lost ground here. A fuller range likely gives visitors a better fit and a stronger price anchor, so keep the extra tier rather than trimming it away.

Targeting / Triggering

Refine where and when the test fires, beyond the page types chosen above in order to increase power (and shorten the testing duration).

Page-level

Enroll anyone who lands on a scoped page.

Targeting notes optional

Interaction-based

Enroll only after a specific action (e.g. add-to-cart, click) to increase power.

Metrics & Ship Requirements

Mark each metric as Primary, Secondary, or Guardrail, then set the requirement it must meet to ship.

Metric	Primary	Secondary	Guardrail	Superiority	Non-Inferiority	Binary	Continuous
Revenue					%
Add to carts					%
Sales					%
Plan 1 Chosen					%
Plan 2 Chosen					%
Plan 3 Chosen					%
Plan 1 Purchased					%
Plan 2 Purchased					%
Plan 3 Purchased					%
Revenue (LTV-implied)					%
Cancellations					%

Power, Sample & Duration

The effect you want to detect and how long you'll run to reach the power for it.

Baseline

MDE

Power

Significance

Daily Traffic

Test type

Total Sample — visitors across all variants

Duration —weeks

Meta-Analysis

Combine your experiment with that of others to reach a decision faster or increase power.

Combine With Other Tests To

Increase Power Decrease Sample or Duration Both

Start Rules

When can the experiment start?

Anytime

Start as soon as it's ready - maximum velocity.

After Test X ends

Wait for another test to finish before starting.

Together with Test X

Run concurrently with another test.

Specific date

Start on a fixed calendar date.

Stop Criteria

When to end the experiment. Pick both for a hybrid.

Fixed Horizon

Run to the planned end, then decide once.

Stop at

Sample sizeDuration

Sequential

Stop early once the result is clear.

Segments

Which subgroups should we track and break the results down by, to see if the effect differs across them?

Segment	Exclude	Measure
Active Customers
New vs. Returning Users
Active Free Users Who Hit a Usage Limit

Iteration Rules

What next experiments to run once the results come in.

Outcome	→ Re-run with: Higher intensity	Higher power	Elsewhere	Something different	Invert	Move on
IF Stat sig primary A clear, significant win.
IF Stat sig primary Flat secondary A win on primary, no effect on secondary.
IF Flat primary Stat sig secondary No effect on primary, a win on secondary.
IF Flat primary Flat secondary No significant effect on either.
IF Stat sig neg primary A clear, significant loss.

Minimum Iteration Attempts

Now Run This Experiment

Do It Yourself with Evidence with access to similar A/B tests Starting at $72/month(with any GoodUI plan)

Unlock the evidence behind each variable
All evidence-backed defaults revealed

Teach Me or My Team with an experiment review Starting at $4,900

Design review
Power and metrics review
Targeting and scope review

Do It For Me with the test designed and built for you Starting at $9,500

Full implementation and QA
Test run and monitoring
Result review and write-up

Beyond the review, we teach you or your team the reasoning behind it, so the next test is one you can design and judge yourselves. Pick whichever lessons you're interested in:

Reading A/B test results How to interpret significance, confidence intervals and what they do and don't tell you.
Avoiding common testing traps Peeking, underpowered tests, metric soup and other pitfalls that quietly bias your reads.
Choosing the right metrics Picking a primary metric and guardrails that match the decision you're actually making.
Sizing and powering a test Baseline, MDE, power and runtime — how to know whether a test can be decisive before you launch.
From result to next experiment Turning an outcome into the next hypothesis: what to keep, change or retest.

Hand it off. We design, implement, and QA the test, run and monitor it for you, then review the results and deliver a write-up — you get the outcome without the build effort.

Metric alignment and setup Define the primary metric, guardrails, target audience, tracking events, attribution window and launch criteria so the test measures the decision you actually need to make.
Power calc Estimate baseline conversion, minimum detectable effect, sample size and runtime before launch, with a clear read on whether the test is likely to be decisive or exploratory.
Prototype-quality code Build the variation to prototype quality in your testing tool or front-end stack, with responsive states, analytics hooks, QA notes and handoff-friendly implementation details.
Analysis Read the result with significance, confidence intervals, guardrail checks and segment caveats, then translate the stats into a practical ship, stop or iterate recommendation.
Post-test and iteration recommendations Turn the result into next moves: what to keep, what to change, what to retest and which follow-up experiments are most worth running next.
Discussions Use async or live discussion time to clarify assumptions before launch, review tradeoffs during the run and unpack ambiguous outcomes after the test closes.

Experiment Design Review +$1,000 optional upgrade One written review of this test design before launch, then a 1-hour call to go through it together, covering:

Design
Power & MDE
Metrics & guardrails
Targeting & scope
Iteration suggestions

Run This Test