Does Charm Pricing Actually Work? Real A/B Test Data on 49 vs 50 Endings

A three-location cafe chain in Izmir Bornova decided to finally settle a marketing argument in autumn 2026: does a 49 price tag really outsell a 50 tag? Six weeks and 18,400 orders later, the data showed something subtler than the textbook answer — and a clear case against blanket charm pricing.

The Experiment Setup

Twelve identical items — lattes, cheesecake, sandwiches — were split across the three branches. Branch 1 used 49₺ tags, Branch 2 used 50₺, and Branch 3 acted as control. After two weeks, the chain rotated the tags between branches to neutralize location effects. They tracked order volume, basket size, and net revenue daily.

The hypothesis seemed boringly obvious: 49 wins, period. Reality was more interesting — charm pricing won on volume, lost on basket, and net revenue depended on the category.

Volume vs Basket Size

The 49₺ tag drove 3.8% more orders versus 50₺. Classic charm pricing effect, confirmed. But the average basket on the 50₺ side was 6% larger — customers who saw a round price seemed to mentally re-anchor and add a second item more readily.

The two effects nearly cancel out, but not quite. Net revenue: the 50₺ tag won by about 2.1% on gross sales. Fewer transactions, bigger tickets.

Category-by-Category Verdict

Charm pricing is not a one-size answer. The data sorted neatly into three buckets:

• Coffee and snacks (impulse buys): 49₺ won. Volume effect dominates, premium signaling irrelevant.
• Full meals and platters: Round prices (50₺, 100₺) won. Customers perceive higher quality and add sides.
• Combos (drink + dessert): 49 + 49 = 98 felt cheaper than 50 + 50 = 100 only when totals were visible — when bundled as a single 98₺ line, it actually reduced attach rate.

One more surprise: the same 49 tag worked better in Turkish lira than in euro pricing previously tested in a partner shop — the bigger absolute number ("249") amplifies the digit-cutoff effect more than small "9" or "19" values.

FAQ

Does charm pricing always beat round numbers? No. Volume rises about 4%, but basket size drops about 6%. Net revenue often favors round numbers, especially on full meals.

When should I use 49 endings? On impulse items: single coffees, pastries, single-serve drinks. Skip charm pricing on premium dishes and bundles.

How do I test this on my own menu? Run a four-week minimum A/B with rotated branches or weeks. thMenu price history exports let you measure volume and basket deltas without manual spreadsheets.

Found this helpful? Share it.