Framing Effects
Date: 2026-08-17
Logically identical information produces different decisions depending on how it’s presented. There is no neutral framing available — you are always choosing one — so the honest question is which true frame to use, not whether to frame.
A framing effect occurs when the same facts, expressed differently, lead to different choices. The classic demonstration contrasts gains and losses:
"90% fat free" → bought more often
"contains 10% fat" → bought less often
identical product
"95% of customers → reassuring
had no problems"
"1 in 20 customers → alarming
had a problem" identical rate
The main frames
GAIN vs LOSS
"save £10" vs "don't pay £10 extra"
→ losses weigh heavier — Loss Aversion
ATTRIBUTE
"90% lean" vs "10% fat"
GOAL
"get whiter teeth" vs
"avoid yellow teeth"
TEMPORAL
"£120 a year" vs "£10 a month"
vs "33p a day"
→ identical cost, very different
perceived size
REFERENCE POINT
"£10 off" vs "£40 instead of £50"
→ Anchoring
See: Loss Aversion · Anchoring
Temporal reframing is the most-used in commerce and the most straightforwardly honest of them, provided the payment option it describes actually exists.
Absolute versus relative
The one worth being careful about, because it’s where framing shades into misleading:
DISCOUNT ON £20 DISCOUNT ON £500
"25% off" ✓ "£125 off" ✓
"£5 off" ✗ small "25% off" ✗ vaguer
people respond to the LARGER-SOUNDING
number, which depends on the base
Both are true. Choosing the one that sounds bigger is normal practice and stays honest as long as the underlying figure is real and available. It stops being honest when the base price is inflated to make the percentage work — Price Anchoring.
Where it lands in an interface
- Delivery — “free delivery over £40” versus “£4.95 delivery under £40”. Same policy; the first frames it as an achievable gain — Shipping Thresholds
- Stock — “only 3 left” versus “3 in stock”. Same fact, one implies scarcity — Scarcity and Urgency
- Returns — “30-day returns” versus “return within 30 days”. The first is a benefit, the second a deadline
- Guarantees — “money-back guarantee” versus “refund policy” — Risk Reversal
- Form fields — “optional” versus “required” labelling changes completion, and marking the smaller set is clearer
- Errors — “password must be 8+ characters” beats “invalid password” because it frames the requirement rather than the failure — Error Prevention and Recovery
There is no neutral frame
Every presentation is a frame. “£4.95 delivery” and “free delivery over £40” and “delivery from £0” are all framings of one policy, and refusing to choose means defaulting to whichever the platform emits.
Which makes the useful test:
1 is it TRUE?
2 does it help the person decide well?
3 would they feel misled on discovering
the other framing?
Question 3 is the honest one. A customer who learns the “free delivery” threshold exists feels informed. A customer who learns the “was” price was never charged feels deceived — Deceptive Design.
Where it’s oversold
- Effect sizes vary enormously by context and by how much the person cares. Framing moves marginal decisions; it rarely moves committed ones
- It doesn’t survive comparison. A customer with two tabs open sees both frames, and the framing advantage disappears
- Repeated exposure erodes it. Regular customers learn what your framings mean
- It’s a presentation lever, not a product lever. Framing a bad offer well produces a return rather than a sale
Testing it
Among the cheapest and safest things to test — no price changes, no engineering, and it’s genuinely reversible.
Test the framing of delivery thresholds, stock messaging, guarantees and discount expression. Measure on contribution and returns, not conversion alone — a frame that converts by overselling shows up later — A-B Tests, Guardrail Metrics.