# Charm pricing: definition, the evidence and the break-even math for $29 vs $30

Source: https://leverprice.com/glossary/charm-pricing
Updated: 2026-11-17

> Charm pricing means ending a price just below a round number, such as $29 instead of $30 or $9.99 instead of $10. Field experiments found prices ending in 9 increased demand, most for new items. The math is forgiving: $30 earns more than $29 as long as it keeps at least 96.7% of $29's buyers.

Charm pricing is pricing a dollar or a cent below a round number: $29 rather than $30, or $9.99 rather than $10. Example: a SaaS plan at $29 a month is charm-priced; the same plan at $30 is not.

## How it works

Buyers read prices from the left and give the first digit extra weight. Thomas and Morwitz found that a nine-ending price is judged smaller than a price one cent higher only when the leftmost digit changes, as in $2.99 against $3.00 ([Journal of Consumer Research](https://doi.org/10.1086/429600)).

Sales data backs this up. Anderson and Simester ran three field experiments that manipulated price endings, and a $9 ending increased demand in all three, more so for new items than for items the retailer had sold before ([Quantitative Marketing and Economics](https://doi.org/10.1023/A:1023581927405)). Their reading: 9-endings help most when customers have little price information.

The revenue math for the higher round price:

```text
$30 wins if conversion at $30 ÷ conversion at $29 > 29 ÷ 30 = 0.967
$100 wins if conversion at $100 ÷ conversion at $99 > 0.99
```

If $29 converts 3.0% of visitors, $30 has to convert at least 2.9% to earn the same.

## Why it matters for pricing

The charm effect and the price level are tangled together. Moving from $29 to $30 is both a lost charm ending and a 3.4% price increase, and only a test separates them. Compare $29, $30 and a clearly higher price such as $35 on [revenue per visitor](https://leverprice.com/glossary/revenue-per-visitor); the [$29 vs $30 breakdown](https://leverprice.com/blog/charm-pricing-29-vs-30) goes deeper.

Endings are also local. A converted US price rarely lands on a familiar number, so Leverprice's regional prices round each country's price to amounts people are used to seeing there, such as 11.99, 1,480 or 999.

## Common mistakes

- Assuming the 9-ending always wins. The field evidence comes from catalog retail, not from software sold to businesses.
- Testing $29 against $30 with too little traffic. A gap that small needs a large [sample size](https://leverprice.com/glossary/sample-size) to detect.
- Converting $29 into another currency and showing the raw result, such as 26.73.
- Changing the ending and the plan's features in the same test.

## Related terms

[Price anchoring](https://leverprice.com/glossary/price-anchoring) is the other big effect of how numbers are presented. [Price elasticity](https://leverprice.com/glossary/price-elasticity) decides whether a 3.4% increase loses more buyers than it gains in revenue, and [conversion rate](https://leverprice.com/glossary/conversion-rate) is where a charm effect would show first.

## Frequently asked questions

### Does charm pricing work for B2B SaaS?

The published field evidence comes from consumer retail. Test it on your own traffic rather than assuming it carries over to software that businesses buy.

### Is $29 or $30 better?

Whichever earns more revenue per visitor on your traffic. $30 only has to keep 96.7% of $29's buyers to come out ahead.

### Does $28.99 beat $29?

Both start with the same left digit, so the left-digit effect gives the extra cent off little to work with. A test could tell, but a gap that small needs a great deal of traffic to measure.
