Short answer
Treat nine-ending prices as a testable hypothesis, especially at a left-digit boundary. Judge the result by contribution per visit, not conversion alone.
The answer in detail
A price ending in nine may help when it changes the leftmost digit, as in $49 instead of $50. Research also suggests differences between new products, deliberative purchases, gifts, and prestige goods. Treat these as hypotheses to test against contribution per visit, not universal pricing laws.
The benchmark
The source draws on a small research set, mostly US studies using dollars. The USD examples below are fictional illustrations of those mechanisms, not new experiments or currency conversions.
| Study | Finding in its research context |
|---|---|
| Thomas and Morwitz, 2005 | Left-digit changes affected perceived price, such as 2.99 versus 3.00 |
| Anderson and Simester, 2003 | Nine endings increased demand in three field experiments, more for new products |
| Stiving, 2000 | Higher-quality products had rounder prices, especially when quality was difficult to judge |
| Wadhwa and Zhang, 2015 | Round prices suited feeling-led choices; precise prices suited deliberative choices |
| Anderson and Simester, 2001 | Sale-sign effectiveness weakened when many products carried the signal |
The well-known Ariely, Loewenstein, and Prelec anchoring experiment from 2003 found MBA students with above-median final Social Security digits stated acceptable prices 57–107% higher than others. A 2012 replication found only a very weak effect. Neither is a forecast of an online store's response, and the cited studies do not measure your contribution margin.
The rule
Start by testing a meaningful left-digit threshold: $49 versus $50, or $199 versus $200. Cutting $48 to $47 costs a dollar per sale without invoking that particular mechanism. That does not prove it can never affect demand; it means the left-digit evidence does not support the same claim.
Round prices may fit products whose value depends on perceived quality or emotion, such as designer pieces and gifts. Widespread .99 endings may conflict with premium positioning. The sale-sign research also suggests that price signals can lose distinctiveness when used everywhere; applying that finding to an entire price-ending system remains an inference.
Use only genuine comparison prices. Anchoring is no justification for invented crossed-out prices; see the reference-pricing guide. The source's lowest-price-in-thirty-days framing concerns European/French promotional rules, not all countries.
The steps
- Audit candidates. Find products just above meaningful boundaries, such as $50–$52 or $100–$105. Rebuild that list in the actual selling currency; do not mechanically convert another market's prices.
- Check positioning. Consider testing nine endings on entry products, consumables, or unfamiliar new items. Retain rounded prices as a hypothesis for emotion-led and prestige categories.
- Check genuine anchors and coherence. Place an honestly priced higher-specification option beside the target. A niche-perfume audit found an eleven-sample set at €41 while single samples cost €1–€1.50. The mismatch undermined the set's value story. It did not require a psychological trick to fix.
- Calculate break-even. Use amounts net of applicable tax:
required volume increase = current unit contribution ÷ contribution at new price − 1. In the source's fictional 20%-VAT example, $50 including VAT is $41.67 excluding VAT. With $20 cost, contribution is $21.67. At $49 including VAT it becomes $20.83, requiring about 4% more sales. Substitute actual destination tax treatment and costs; 20% is only this illustration. - Test in context. Compare changed products with a similar unchanged group over at least two buying cycles. Each shopper should encounter consistent prices. One Club winter-accessories brand excluded already-discounted products to avoid reference-price complications and chose contribution per visitor as the outcome, building on revenue per visit. With low volume, a dated before-and-after analysis can be directional, with weaker causal confidence.
The pitfall
Moving the entire catalog to .99 because “it works” sacrifices both margin and useful comparison. A lower price may improve conversion and still reduce contribution. Nor does lower pricing guarantee conversion improvement in every setting.
A nine ending can also break a cart threshold. While preparing a test, a low-basket baby-products brand found that repricing an accessory to €8.90 made a common combination total €38.80, just below free shipping at €39. Recalculate common baskets against your shipping threshold before lowering prices. Keep the original euros as case evidence rather than a recommended global price ladder.
What to tell your team
These figures are fictional:
“We have twelve products at $50–$52. We propose testing $49 for six weeks against a comparable unchanged group, judging contribution per visit. Our illustrative cost structure requires about 4% more sales to break even. Gift sets will keep round prices.”
Sources
- Thomas and Morwitz: Penny Wise and Pound Foolish, Journal of Consumer Research, 2005
- Anderson and Simester: Effects of $9 Price Endings on Retail Sales, 2003
- Stiving: Price-Endings When Prices Signal Quality, Management Science, 2000
- Wadhwa and Zhang: This Number Just Feels Right, Journal of Consumer Research, 2015
- Anderson and Simester: Are Sale Signs Less Effective When More Products Have Them?, 2001
- Ariely, Loewenstein and Prelec: Coherent Arbitrariness, 2003
- Fudenberg, Levine and Maniadis: On the Robustness of Anchoring Effects, 2012
