How to set, raise and localize prices

1 The bottom line

Most apps charge too little, so find the price users call fair, test higher prices on a slice of users, and adapt by country. Prices can be rolled back, so testing costs little.

2 Key takeaways

• Ask users about price in relative terms, not "what should I charge". People give weak answers to open price questions but good answers when they compare.
  • The median answer to "starting to get expensive" was $30 a month at Superhuman, which set its price Rahul Vohra, Superhuman1:07:45.
  • Ladder's research showed users compared it to Spotify and Netflix, so it cut from $60 to $30 a month Greg Stewart, Ladder13:30.
• Test higher prices, because demand is often less sensitive than founders fear. A higher price can signal value and bring in buyers who value the product more.
  • Glam Up went from $3 to $9 a week and conversion rose Nicole (madebycol), Glam Up and Sprout23:00.
  • Coconote tested $129 a year against $99 and got more users and more revenue Brett Bauman and Zack Hargett, Coconote40:15.
• Run price tests on a slice of users and keep raising until the numbers break. A bad test only touches a few users and can be undone.
  • Test on 10, 15 or 20% of the audience, not everyone Josh, Blue Throne33:45.
  • If a test fails, you can roll back and nobody will notice Jacob Eiting and David Barnard, RevenueCat42:00.
• Localize price by market instead of converting currency. Willingness to pay differs by country, and exchange rates ignore that.
  • Lower prices in Brazil and Mexico beat US-equivalent prices, as part of a 60% ARPU lift Michal Parizek, Mojo4:59.
  • Price cuts of 20 to 30% in some markets caused a surge in new subscribers Zain Abiddin, Canva14:15.

I learned your price is never final.

Melissa Cash, Pok Pok

2.1 Find a price users feel is fair

Ask users to compare, and set a range from the value of the job and the cost of alternatives.

• Ask about 100 early users the Van Westendorp questions. Source: Rahul Vohra, Superhuman1:07:45
  • Ask four things: too expensive to consider, so cheap you would doubt quality, starting to get expensive, and a bargain.
  • Most startups anchor to the bargain price, but a best-in-category position is supported by the "starting to get expensive" price.
  • At 10x ARR, a $1 billion valuation needs 300,000 subscribers at $30, so gut-check market size against your goal Rahul Vohra, Superhuman1:07:45.
  • OpenAI had no time for a pricing expert, so it used a Discord, a Google Form and a blog post on the method, which led to $20 a month Nick Turley, OpenAI40:47.
  • Gamma used Van Westendorp and conjoint-style surveys of early users to land on a single plan of about $20 a month Grant Lee, Gamma1:30:15.
• Ask price questions in relative terms. Source: Madhavan Ramanujam, Simon-Kucher25:00
  • Asking customers what to charge gives garbage answers.
  • Instead, index a known product at 100 in value and ask where you sit, then ask where your price should sit if it is 100 in price.
  • A 110 answer means you can price above the benchmark Madhavan Ramanujam, Simon-Kucher25:00.
  • Ask what is fair, what is expensive and what is prohibitively expensive. The fair price is often a bargaining answer, while the expensive price is what they would pay if the product is good Todd Jackson, First Round Capital1:21:17.
  • dbt Labs compared its worth to customers' cloud warehouse spend and talked to dozens of customers before its first pricing change Julia Schottenstein, dbt Labs36:06.
• Bound your price by the value of the job and what customers spend on alternatives. Source: Rob Snyder55:50
  • If market research consultants cost $50,000 a year, charging about the same feels intuitive to the buyer.
  • Early on you are testing willingness to transact, not willingness to pay.
  • Heavy Tinder spenders compared the cost to dating, a couple hundred dollars a month, not to other subscriptions. That let Tinder price the Super Like much higher than first expected Ravi Mehta, Outpace37:00.
• Let beta customers choose what they pay. Source: Karri Saarinen, Linear1:15:03
  • Linear added a settings page where customers could optionally pick a per-seat price up to about $20.
  • Some paid $28 or $1, which was not enough data to set a price, but it tested the payment flow and gave confidence that a price under $20 would find a market Karri Saarinen, Linear1:15:03.
  • At launch, nearly all several hundred beta companies subscribed to the paid plan Karri Saarinen, Linear1:15:03.

2.2 Show the price in a way users can judge

One clear number, framed as a small recurring amount, beats ranges and big totals.

• Show one price, not a range. Source: Uber
  • Users fixated on the high end of Uber's $14 to $20 range, which created uncertainty.
  • A single estimate led to a double-digit increase in rides per user Uber.
• Make the monthly price the headline number. Source: Unacademy
  • A smaller recurring amount is easier to judge than a large lump sum.
  • The monthly variant produced 13.5% more click-through rate Unacademy.
• Push the annual plan, and test a higher annual price. Source: Brett Bauman and Zack Hargett, Coconote40:15
  • Coconote launched at $99 a year and $19.99 a month.
  • About 80% of subscribers took the annual tier Brett Bauman and Zack Hargett, Coconote40:15.
• Pick an intro price that feels premium but safe to try. Source: Steve P. Young, App Masters6:43
  • Raising a paid kids app from 99 cents to $1.99 lost customers but earned more money.
• Give price-sensitive users a discount, and let them pick it. Source: Steve P. Young and Rosie, App Masters0:43
  • Do not drop the base price for everyone. Some users will never pay the first price they see, so give those users a discount offer.
  • Show up to three options: the one you want them to buy, a cheaper one and a pricier one Steve P. Young and Chaim, B7 Dev0:16.
  • Letting users pick the discount beat a flat 50% off in A/B tests, and some users even pick the higher price Steve P. Young, App Masters5:26.

2.3 Test higher prices safely

Raise price on a slice of users, keep going until demand breaks, and know you can undo it.

• Push the price up on 10 to 20% of users until the numbers break. Source: Josh, Blue Throne33:45
  • Do this even if you doubt your app can command $20 a month.
  • At $4.99 a month you need about half a million payers to reach $30M ARR. At $20 you need only 125,000 Josh, Blue Throne33:45.
  • Waterspeed at roughly 50 cents a month was projected at about $500 a month, so low prices cap the business Massimiliano Picchi and Steve P. Young, Waterspeed7:53.
• Expect conversion to hold or even rise when you charge more. Source: Nicole (madebycol), Glam Up and Sprout23:00
  • Her theory is that demand is inelastic for apps people really want, and a low price can signal low value.
  • On her newest app she raised both plans by 20%, and conversion and revenue per download rose. She is testing another 20%.
  • Moving from $20 to $30 in the US raised both trial conversions and revenue Steve P. Young and Vitaly Davydov, App Masters13:27.
  • Gamma tested $8, $10, $12 and $20, and $20 won on revenue and conversion Jon Noronha, Gamma19:31.
  • Moving $5.99 to $7.99 rarely hurts much, because roughly 20% of any audience will pay a lot more Idan Waller and Steve P. Young, BlueThrone25:20.
• Fix price and paywall before building features. Source: Steve P. Young and Chaim, B7 Dev7:45
  • The app already ranked for keywords, so it had organic downloads to convert better.
  • The change was a hard paywall and $60 a year, up from a one-time $2 to $3 Steve Young, App Masters18:35.
  • Revenue went from $1,000 to $2,000 a month to $8,000 to $10,000 a month, with no new features Steve P. Young and Chaim, B7 Dev7:45.
  • The app got 100 to 150 downloads a day Steve Young, App Masters18:35.
• Tie a price rise to a visible jump in value or proof. Source: Sameer Ahuja, GameChanger39:50
  • Before asking users to pay for live streaming, GameChanger upgraded video to 1080 HD and added highlight reels, lineup recommendations and more.
  • Steve announced his Academy would rise from $25 to $49 a month on a set date, which gave buyers a reason to sign up now Steve P. Young, App Masters7:54.
• Do not undercharge out of fear. Source: Jacob Eiting and David Barnard, RevenueCat42:00
  • Founders often undercharge because of impostor syndrome and fear of upsetting users.
  • Pricing is transient, so a failed price or hard paywall test can be rolled back.
  • Each pricing option still needs development work and paywall logic. Mindful Mamas started with just two products Jillian Lee Stout, Melody Boselin and Sunny Dulay, Mindful Mamas30:11.

2.4 Read the results and settle on a price

Choose the price customers feel is fair, find out why a price fails before cutting it, and keep checking.

• Pick the price customers feel is fair, not only the one that earns most. Source: Michael Segal and Mark Ungerer, Skylight51:12
  • Combine interviews with quant tests, since people sometimes say one thing and do another.
  • The spreadsheet said $99 would earn a little more than $79, but interviews put $99 near "disgust" territory, so they shipped $79 Michael Segal and Mark Ungerer, Skylight51:12.
  • Going from $39 to $79 a year cost minimal retention Michael Segal and Mark Ungerer, Skylight51:12.
• If a price rise changes nothing, raise it again. Source: Jason Cohen, WP Engine38:57
  • Pricing selects your market. Larger buyers read a very low price as low quality and will not buy.
  • A founder moved from $300 a year to $300 a month and still got one or two signups a week Jason Cohen, WP Engine38:57.
• Find out why a price is too high before cutting it. Source: Greg Stewart, Ladder13:30
  • Users did not want or value the coach chat that justified $60, so Ladder cut the chat and landed at $30.
  • When reviews said it was too expensive, AllTrails tested $50 against $30 and $15 on the web. $15 and $30 behaved similarly, so it chose $30 Ron Schneidermann, AllTrails18:19.
  • Cheaper will not fix retention if people do not care about the product Jack McDermott and Tanmay Jain, ElevenLabs58:35.
• Verify after a raise and keep watching retention. Source: Miranda Paine, Runna31:44
  • Runna benchmarked similar apps, tested $19.99 monthly and $119.99 annual, then ran willingness-to-pay research that showed the price was about right.
  • Revenue, trial starts and conversion all rose together Miranda Paine, Runna31:44.
  • Pok Pok doubled its price and doubled revenue, then lowered it a bit after seeing longer-term retention Melissa Cash, Pok Pok42:45.

2.5 Adapt pricing by country

Find where your revenue comes from, then test local prices instead of converting the US price.

• Check which countries send downloads and revenue. Source: Steve P. Young, Learna1:59
  • Use AppFigures to see top countries for competitor apps.
  • Learna gets only about 20% of downloads from the US. Mexico is number two, with Thailand and Vietnam also high Steve P. Young, Learna1:59.
  • Routinery's top countries are the US, South Korea and Japan, markets with real spending power TP Young, Routinery1:27.
• Test local prices instead of exchange rates, and expect cuts in some markets. Source: Zain Abiddin, Canva14:15
  • Of local currency, local price points and local payment methods, price adjustment had the biggest impact.
  • Mojo first used App Store style conversions, then found lower prices in Brazil and Mexico performed better Michal Parizek, Mojo4:59.
  • ARPU tracks per capita income, so you cannot extract $100 per user where average income is about $2,500 a year Kunal Shah, CRED24:30.
• Build one pricing index for all countries, roll it out as a test, and keep adjusting. Source: Yoav Sharon, Simply51:36
  • Simply made its own Big Mac style index that combines purchasing power with how people value learning an instrument or drawing.
  • ARPU rose dramatically in most geos, but a few did not work and needed adapting Yoav Sharon, Simply51:36.
  • Tinder's fastest growing markets, Latin America and India, had very different willingness to pay. Start in core markets and expand as the US saturates Jeff Morris, Tinder20:50.
• Show local currency and offer a local payment method. Source: Steve Young and Lucas, Paddle29:43
  • Use methods such as Alipay in China or Pix in Brazil. Local currency removes exchange rate uncertainty.
  • Companies with at least one alternative payment method grow 20 to 25% faster. Even two currencies gave 12.7% Steve Young and Lucas, Paddle29:43.
• Segment tests by US and international users. Source: Morrie Schonfeld, Pingo48:10
  • The split showed 3-day and 7-day trials perform very differently by region.
  • They were roughly equal in the US, but 7-day was about 2% higher in conversion internationally Morrie Schonfeld, Pingo48:10.
  • Expect Android to convert about half as well as iOS, sometimes less Josh and Steve P Young, Miracle Morning20:10.

2.6 Where they disagree

Whether to price low or high depends on your audience, your goal and your stage.

• Low price as a strategy, or higher price for revenue. Fit the price to your market, not to another app.
  • Cal AI stayed at $30 a year while MyFitnessPal charged about $70, optimizing for mind share, not profit Jake Castillo, Cal AI56:10.
  • A low price feels like an impulse purchase, so people share the app freely Daniel Heintzman, Cal AI33:10.
  • Sunflower's users often have little money, and it made more gains by lowering its price than raising it Koby Conrad, Sunflower46:00.
  • DoorDash kept DashPass at $9.99 a month since 2018, because a $1 to $2 rise would lift revenue 10 to 20% but slow growth over 5 to 10 years Jimmy Liu, DoorDash34:10.
• Start high and lower later, or start low as a test and raise.
  • Lowering a price is welcome, and raising it later makes customers unhappy Chaim Sajnovsky and Steve P. Young, App Masters16:38.
  • Superpower says to launch at the price you want long term, because starting at $10,000 led to decisions for that price point, and it cut to $499 a year Max Marchione, Superpower8:16.
  • Steve launched his Academy at $25 a month as a test to collect case studies, then raised it to $49 Steve P. Young, App Masters7:54.
  • Pok Pok launched at $29.99 a year out of fear, and doubling it doubled revenue Melissa Cash, Pok Pok42:45.
• Discounts help low-risk buyers but not risk-bearing ones.
  • Discounting does not reduce risk, so before the chasm use value pricing and a guarantee instead. Discounts and freemium fit only when adoption risk is low Geoffrey Moore1:03:40.
  • For consumer apps, a user-picked discount beat a flat 50% off Steve P. Young and Chaim, B7 Dev0:16.
• Cut local prices, or match the number.
  • Canva and Mojo saw gains from lower local prices.
  • In Australia, a $40 US price became 40 Australian dollars, a nominal match that is cheaper in US dollar terms than a straight conversion, with immediate success Steve P. Young, App Masters13:47.