How Big Macs Increased Our Revenue by 15%

A (surprising?) case study of geographical pricing

Yuval Kaminka
Oct 2, 2017 · 5 min read

TL;DR an experimental pricing scheme for global sales increased our music learning revenue at JoyTunes by 15%. Yes, it was based on Big Macs.

“You know what made us the biggest, meanest, Big Mac eating, calorie-counting, world-dominating kick-ass powerhouse country in the history of the human race? The pursuit of happiness. Not happiness. The pursuit.”

— Will Ferguson

Quick Background

Our ambitious mission at JoyTunes is to make it easy, fun and affordable to learn to play any musical instrument. Currently, our super fun and engaging piano learning apps are seeing massive growth, due in no small part to our numerous product improvements and tests.

Here, I focus on one such test, specifically for pricing our memberships. The motivation came from a boost in sales that we’ve seen with our China users when we hit the magical 50% discount.

Our piano learning apps have always gained a lot of quality users in China, but the conversion to memberships was much lower than in the US, which was frustrating. We knew things were generally cheaper in China and had some ideas for the right pricing. And so, we figured we could try to gradually lower the prices and see what happens.

Lo and behold, as the price went down, the conversion went up. So far so good. But the conversion increase wasn’t enough to compensate for the reduced price.

Let’s see what this means. In broad strokes, the revenue is the multiplication of the number of users, the conversion and the price —

Rev = N_Users × Conversion × Price

If we lower the price by 25%, for example, then we’ll have a new conversion and therefore new revenue (we’ll safely assume the number of users remains the same) —

Rev_2 = N_Users × Conversion_2 × Price × 0.75

If we want the revenue to increase, meaning Rev_2 > Rev, then it turns out in the example above that Conversion_2 needs to be at least 33% higher than the initial conversion. Anything below means you’re losing money. Which is exactly what happened to us, that is, until we hit that magical 50% discount.

Remember, according to the equations above, or just plain logic, we need to double the purchases (meaning double the conversion) to compensate for cutting the price by half, but we more than doubled. Way more.

Yada Yada Yada, we’re doing a geographical pricing test…

Big Mac Pricing Scheme

The China success led to the desire of finding a pricing optimum for our long-tail downloads outside the US and China, that account for roughly 40% of our users. As in many cases, we had to hack our way through this as we don’t have the capacity to carefully price each country separately. Besides, most countries would need an abundance of time to accumulate enough data for a reasonable outcome (more on that in a separate post). And so, we used the geographical bulk pricing options to create an overall scheme.

In short, we bulk-adjusted the membership prices according to the Big Mac Index.

The Big Mac Index is a surprisingly insightful pseudo financial index created by the Economist. The index uses McDonald’s Big Mac prices around the world to compare and measure currency ‘strength’. At its core, the index is basically just a list of Big Mac prices in over 50 countries as compared to the US prices. Interestingly, it’s helpful in figuring out the relative purchasing power when it comes to mainstream commodities, that is, assuming McDonald’s have been doing their homework (and they have!).

As our mission is to make it possible for everyone to learn an instrument, and as iOS subscriptions are still fairly new, we wanted to make sure we fit mainstream prices.

“A Big Mac — the communion wafer of consumption.”
— John Ralston Saul

  1. Iteration I: We adjusted all the prices according to the Big Mac Index, meaning, using US prices as a base, we decreased or increased (mostly decreased) the membership prices based on what the Big Mac index was for that country. We did this for 60 countries.
    (See below an example spreadsheet)
  2. Sanity: Interestingly, this method led to the pricing we manually found to be the best for China. A quick sanity for some other countries showed quite reasonable pricing. At this point we couldn’t keep still due to overwhelming anticipation.
  3. Iteration II: Taking the result of #1, we applied cosmetic changes for the local currency. For example, a price of 10.99 euros was changed to 9.99 etc.

Use the spreadsheet below to play around with the price adjustments. We used a similar one in our work.

Important Notes:

  1. The Big Mac index changes from time to time, the one in the sheet is from late 2016.
  2. Not all countries appear in the index. For those that don’t you can set a default adjustment. Since these accounted for a negligible share of sales we just kept it at 100% of the US price.
  3. Some prices are not in USD. I adjusted from USD to the local currency using a currency conversion that is relevant for late 2016. This changes from time to time.

The worldwide revenue increased by 15%!!!
(yes, it was statistically significant)

Not bad for a day’s work :)

Alternative Approaches

-or- even better ideas!

There are obviously many alternatives to our choice of global pricing. These are a just few that come to mind —

  • Unified Pricing — iPads and iPhones cost roughly the same worldwide (up to varying taxes), why should memberships be any different? Moreover, affording an iPhone might make one less sensitive to a-few-dollars-per-month difference in education services(?)
  • Use some other scheme — For example, adjusting the membership prices to piano lesson prices worldwide. We played around with this notion, also using askwonder for a sample of lesson prices in key countries
  • Individual testing — As with China, there could be other magical singularity pricing points for different countries and cultures. It would be interesting to explore, or better yet, automate the process of optimizing the price over time


We run hundreds of tests of various shapes and sizes, which has led to 10x revenue growth over the past year and many more satisfied piano learners. This was but one of these tests — we hope it was helpful.

“I am the literary equivalent of a Big Mac and fries.”
— Stephen King


Sharing some inner workings from JoyTunes' journey to create the biggest music learning service: iOS & Android development, deep learning, machine learning infrastructure, subscription dynamics, startup culture and more

Yuval Kaminka

Written by

Extremely passionate about working on things that matter. Co-founder of JoyTunes, making it possible for anyone to play a musical instrument



Sharing some inner workings from JoyTunes' journey to create the biggest music learning service: iOS & Android development, deep learning, machine learning infrastructure, subscription dynamics, startup culture and more

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade