Framing Effect

Metricuno
May 17, 2026
4 min read
Framing Effect — How the framing effect changes shopper decisions — with copy, pricing, and discount examples, lift benchmarks, and how to test it on your store.
Quick answer

The framing effect is why "90% fat-free" outsells "10% fat" and "$50 off" can beat "15% off" — a behavioral lens that quietly drives copy and pricing tests.

Definition
Behavioral economics

Framing Effect

A cognitive bias where the same information produces different decisions depending on how it's presented.

The framing effect is the tendency for people to react differently to identical information depending on whether it's framed as a gain or a loss, a percentage or an absolute number, or a positive or negative attribute. A product described as '90% fat-free' feels healthier than the identical product described as '10% fat'. A '$50 off' label on a €200 jacket converts differently than '25% off', even though the discount is the same.

For online stores, it's the single most useful behavioral lens for copy, pricing, and discount experiments — because the underlying offer never has to change.

Also known as
Frame dependence
Presentation bias

Daniel Kahneman and Amos Tversky formalised the framing effect in 1981 with their 'Asian disease problem', showing that subjects flipped their choice between identical outcomes when one was described as 'lives saved' versus 'lives lost'. The mechanism is loss aversion: a loss feels roughly twice as painful as an equivalent gain feels good.

On a product page, that asymmetry shows up everywhere. 'Free shipping on orders over €60' converts differently than '€5.95 shipping under €60'. 'Save €30' reads differently than 'Don't pay €30 extra'. The offer is identical; the frame moves the conversion rate. It's one of the most studied entries under cognitive biases for a reason — it's testable, repeatable, and almost always cheap to ship.

Formula

Framing Lift = (CR_frame_B - CR_frame_A) / CR_frame_A

Variables

CR_frame_A

Control frame conversion rate

Conversion rate of the existing copy or pricing frame.

CR_frame_B

Variant frame conversion rate

Conversion rate of the alternative framing being tested.

Framing Lift

Relative lift

Percentage change in conversion attributable to the framing change alone.

Worked example

A Shopify apparel store tests the discount label on a winter coat. The control reads '20% off'; the variant reads 'Save €40'.

CR_frame_A (20% off): 3.2%

CR_frame_B (Save €40): 3.8%

Framing Lift = (3.8% - 3.2%) / 3.2% = 18.75%

Switching from a percentage frame to an absolute-euro frame produced an 18.75% relative lift in add-to-cart conversion — with no change to price, product, or page layout.

The general rule from the test data: absolute discounts win above roughly €50 (the saving feels bigger as a number), percentages win below that threshold (the percentage feels bigger as a number). Loss-framed urgency ('Don't miss…') typically beats gain-framed urgency on scarcity and cart-abandonment flows. The table below shows the lift ranges teams typically see on Shopify and WooCommerce stores.

Benchmark

Typical conversion lift from common framing swaps on DTC product and checkout pages

Framing patternApparel / accessoriesBeauty / supplementsElectronics / home
Absolute € off vs % off (orders > €50)+12% to +22%+8% to +18%+10% to +25%
% off vs absolute € off (orders < €50)+6% to +14%+5% to +12%+4% to +10%
Gain frame ('Save €X') vs loss frame ('Don't pay €X extra')+3% to +9%+4% to +11%+2% to +7%
'90% positive reviews' vs '4.5/5 stars'+5% to +12%+7% to +15%+6% to +13%
'Free shipping over €60' vs 'Shipping €5.95'+9% to +18%+10% to +20%+7% to +14%
Anchor price shown vs no anchor+8% to +16%+6% to +14%+10% to +22%

Treat these ranges as starting hypotheses, not promises. Framing lifts decay when overused on the same audience, and they interact with category norms — a beauty SKU buyer reads '90% natural ingredients' very differently than a returning customer reads it. The right move is to ship two or three framing variants per quarter on high-traffic pages and let your test platform call the winner.

Frequently asked

Framing effect FAQ

Showing '90% of customers re-order' converts better than '10% don't re-order', even though both statements describe the same data. The positive frame triggers a different decision from the negative frame.

No. Anchoring is about reference points — showing a €200 'was' price next to a €120 'now' price. Framing is about how the same information is described. They often appear together on a product page but they're distinct cognitive biases.

Use whichever number is bigger. On a €200 product, '€40 off' feels larger than '20% off'. On a €15 product, '20% off' feels larger than '€3 off'. The crossover sits around €50-€100 for most categories.

No. Loss framing wins for urgency, scarcity, and cart-abandonment ('Don't lose your 10% off'), but gain framing tends to win for upper-funnel acquisition and brand-positive messaging. Test both per context.

Framing lifts typically run 5-20%, so plan for 8,000-25,000 visitors per variant on a 3% baseline conversion rate. Use a sample size calculator with your real baseline rather than relying on rules of thumb.

It's one of the most testable entries in the cognitive biases family, alongside anchoring, social proof, and the decoy effect. It's especially useful because it requires no change to the underlying offer — only the wording.

Yes, when it actively misleads. '90% fat-free' is honest reframing; '90% chance of success' for a product with no such evidence is deceptive. The bar is whether the literal claim is accurate.

Structure it as: 'Because [audience] responds more strongly to [loss/gain/absolute/percent] frames, changing [element] from [frame A] to [frame B] will lift [metric] by [estimated range].' Pre-register the expected direction.

Start with the highest-traffic, highest-stakes copy: the discount label on the product page, the shipping threshold message in the cart, and the urgency line in cart-abandonment emails. These three together usually account for the biggest measurable wins.

It works in B2B too, but the effect sizes are smaller because B2B buyers are more likely to do explicit comparison and review by committee. In direct-to-consumer purchases, where decisions happen in seconds, framing has more room to move the needle.

Test ideas before you ship them

Run unlimited A/B tests, attach hypotheses to outcomes, and build a searchable archive of what works — and what doesn't.