All answers

The Prevailing
Answers for
Business model:
Business model, pick up to two

How should we test price changes safely?

The PrevailingPublished Oct 28, 2026

Our answer

Run an A/B test on one best seller at prices within 5 to 10% of the current price, and hold each shopper to one random price through checkout. Keep feeds and ads consistent, and watch ad costs alongside profit.

Confidence
Medium
Strongest proof
4 official rules
11 practitioners, 2 opinions or claims
Sources
7 sources
17 statements

New here? Take the 1-minute tour

Why this is the answer

3 reasons and our judgment

Assign each visitor a price at random and hold it through checkout. That is what makes a price test safe, unlike personalized pricing set from data about a person, which draws regulatory scrutiny (Shoplift, Shopify A/B testing app).

Start small, keeping feeds and ads in step. Intelligems (price testing software maker) starts with one best seller at about 5 to 10% either side, keeps ads price-agnostic and sets the Shopify price to the highest test price so feeds never show a lower one.

A test that wins on your site can still lose money on ads. Juan Cruz of MarsMen didn't roll out a test winner with about 25% more revenue per visitor, because Meta cost per purchase rose 50%.

Our judgment

  • Start with one best seller within about 5 to 10% of the current price. That limits how many shoppers and how much revenue a test can touch, going by Intelligems' straddle test and Shoplift's advice to start with two prices.
  • Settle feeds, ads, subscriptions and the rollout plan before launch. Intelligems and Shopify say that is where prices diverge, and Shoplift says that is where complaints come from.
  • Check Meta cost per purchase before launching a site winner. That goes by Juan Cruz's MarsMen test and Intelligems' note that variants with lower conversion can draw higher costs per thousand impressions.
  • The advice to run a big price change and revert if it fails doesn't change the advice to A/B test first. A before and after read mixes the price with season, traffic and promotions, as Intelligems and Booper say, and Common Thread Collective warns that rolling prices back is often the worst move.

Our judgment, built from Intelligems, Shoplift, Shopify, Juan Cruz, Booper, Common Thread Collective and The Andrew Faris Podcast: our reading of these sources, not a finding from any one of them.

What the sources say

17 statements from 7 sources
  • IntelligemsUndatedSupports

    Intelligems, which sells price testing, says it sets Shopify prices to the highest test price, so shoppers who see a lower price in Google Shopping never meet a higher one on the site. End one test before the next.

    “Setting the highest test price in Shopify prevents customers from seeing a lower price in Google Shopping (or other channels) and then encountering a higher price on your site.”

    Read it on docs.intelligems.io

  • IntelligemsUndatedSupports

    Intelligems, which sells price testing, says a price change can be rolled out only once, ending a test needs a sweep for old prices, and subscribers during the test are tied to duplicate products whose renewals fail.

    “A Price Change can only be rolled out once, at the moment you end it.”

    Read it on docs.intelligems.io

  • IntelligemsUndatedSupports

    Intelligems, which sells the testing app, says tests can target by new or returning, device, traffic source and country. It only tracks visits since installation, and source classification is approximate.

    “The source site logic is an approximation that may not perfectly classify all traffic.”

    Read it on docs.intelligems.io

  • IntelligemsUndatedSupports

    Intelligems, which sells a price testing app, says to check each test group on a phone and set non-default currency prices as fixed during a price test, since dynamic conversion can distort results.

    “be sure to set non-default currency prices as fixed/static for the duration of the test”

    Read it on docs.intelligems.io

  • ShopifyUndatedSupports

    Shopify's built-in Experiments need the Grow plan or higher and cannot test headless checkouts or Liquid template changes. During a test, ads and feeds may show a different offer from the store.

    “Product ads, listings, feeds, and third-party apps might not display the same offer.”

    Read it on help.shopify.com

  • ShopifyUndatedSupports

    Shopify Experiments test theme, checkout, catalog or discount changes against a control, by default with a 50/50 traffic split and an end date 90 days after the start.

    “By default, a new Experiment uses a traffic split of 50% control and 50% treatment, and an end date 90 days after the Experiment starts.”

    Read it on help.shopify.com

  • ShopifyUndatedSupports

    Shopify's free Smart Pricing app gives AI price suggestions and lets select US merchants A/B test prices with online customers.

    “Select US merchants can also A/B test prices with online customers.”

    Read it on apps.shopify.com

  • ShopifyOct 23, 2026Supports

    Shopify's Rollouts now handle catalog changes, so stores can schedule product availability and pricing per market and run experiments between catalogs on the online store.

    “Rollouts now supports catalog changes, so you can schedule product availability and pricing per market.”

    Read it on changelog.shopify.com

  • Homestead / Mars MenSep 17, 2026Supports

    MarsMen's ecommerce head says a test that raised revenue per visitor about 25% on site (higher AOV, lower conversion) lifted Meta CPA 50%, so it was not rolled out. He suspects Meta charges more when conversion falls.

    “we ended up seeing an increase in CPA of 50% and then the ROAS stopped making sense”

    Read it on splitbase.com

  • ShopliftJul 31, 2026Supports

    Shoplift, which sells price testing software, says an A/B price test assigns each visitor randomly to one price held through checkout, unlike personalized pricing, which draws regulatory scrutiny. Not legal advice.

    “An A/B price test assigns each visitor at random to one of two prices”

    Read it on shoplift.ai

  • ShopliftJul 31, 2026Supports

    Shoplift, which sells price testing software, says test a bestseller's current price against one higher challenger, then roll the winner out to all traffic and reset the baseline.

    “Start with two, the current price and one challenger.”

    Read it on shoplift.ai

  • Andrew FarisJun 18, 2026Disagrees

    On Andrew Faris's podcast, the speakers say most brands pick prices irrationally and suggest big price tests that narrow over time; one company raised prices 20% with no drag by charging for things it gave free.

    “one of the things that is most irrationally chosen in e-commerce like people just sort of pick a number”

    Read it on podcasters.spotify.com

  • IntelligemsJun 17, 2026Supports

    Intelligems, which sells price testing software, says to test the current price against prices 5 to 10% lower and higher at once for two to three weeks, judging on profit per visitor.

    “Based on data from thousands of price tests, a 5% price change typically moves conversion rate by around 10% in the opposite direction.”

    Read it on intelligems.io

  • BooperMay 28, 2026Supports

    Booper, a pricing software vendor, says measure price elasticity excluding promotions and stockouts, ideally against control stores or with A/B tests. A single price change or a two-week read is not reliable.

    “Elasticity calculated based on a single price change is merely anecdotal, not a meaningful measure.”

    Read it on booper.fr

  • IntelligemsMay 27, 2026Supports

    Intelligems, which sells a testing tool, says a variant that trades conversion rate for order value will show CAC rising, and Meta may charge higher CPMs for it. It gives no data on the CPM effect.

    “any test that trades conversion rate for order value will show CAC rising”

    Read it on intelligems.io

  • Common Thread CollectiveMay 6, 2026Caveat

    Common Thread Collective, an agency, says brands that raised prices for tariffs should not roll back sitewide. Check competitor prices and first-purchase AOV, and compare repeat rates of customers won before and after.

    “If your first-purchase AOV is now above market comparables for your category, you're pricing out the top of the funnel.”

    Read it on commonthreadco.com

  • ShopliftFeb 13, 2026Supports

    Shoplift, which sells a Shopify price testing app, says to treat a launch price as a hypothesis: test around it (say $39, $49 and $59) for two to four weeks in a stable period and judge on revenue per visitor.

    “They tend to overstate what they'd pay when no actual purchase is required.”

    Read it on shoplift.ai

In the full answer

For readers who sign in
  • What to do this week3 steps
  • The numbers behind it3 figures from cited sources
  • The strongest disagreementThe best case against the answer, from a cited source
  • How sure we areWhat its confidence rests on, and what would change the answer

Read next

More in Measurement