The Cupcake Strategy

When designing a new product, make it like a cupcake rather than a wedding cake. With a wedding cake, you master each part (base, frosting, decorations, etc.) but when put together the cake may still taste awful if the parts don’t compliment each other. However,…

When designing a new product, make it like a cupcake rather than a wedding cake. With a wedding cake, you master each part (base, frosting, decorations, etc.) but when put together the cake may still taste awful if the parts don’t compliment each other. However, if you make a cupcake first, the risk of failure is smaller, allowing you to experiment to get it right as you scale up.

So, why was Des Traynor of Intercom using a cake analogy for building software products?  Perhaps he was hungry at the time, who knows! But we actually think the analogy is brilliant.

Des suggests NOT just going out and building the wedding cake from scratch… even if you’ve mastered every part of what goes into baking a wedding cake.  Maybe you’ve learned how to mix the ingredients perfectly… bake the cake… and even decorate it so well that any pastry chef would be proud. If this is the case, why not just get started with the process?

Here’s the thing:  What if you do all of that… but the end result is just bad?  The cake simply tastes awful, despite you having the knowledge and expertise to bake it?  Maybe there was a misstep along the way. Or perhaps the recipe itself just wasn’t good. Either way, you now have an entire wedding cake that nobody wants to eat.

An alternative approach:  

Start with a cupcake.  Bake a much smaller batch of those first, going through the same steps as you would have with the wedding cake.  It should take much less time, but you’ll still be essentially baking a cake, topping it off with frosting, etc. When you bite into that, though, you’ll be able to see if it’s as good as you were hoping.  If it is, then great! Onto one of the layers of the cake. But if it isn’t, you have an opportunity to start over. Perhaps you can change one of the ingredients… or try to tackle a completely different type of cake.  If you do need to change course, you wouldn’t have wasted an entire wedding cake, though. You would have just scrapped a few cupcakes in your quest to making a better wedding cake.

Those involved in building software products already see the analogy.  Rather than approach building a product in a linear form (like a wedding cake), tacking smaller batches of work which you can test, measure, and iterate, if necessary, will likely lead you to a better product.  

So, we certainly appreciate the analogy from Des — and hope that we didn’t make too many Product Managers hungry in the process of our explanation.

Paul McAvinchey
Paul McAvinchey
paul@productcollective.com

For over 15 years, Paul has been building and collaborating on digital products with fast-growing startups and global brands, including AOL and WMS Gaming. Currently, he's a co-founder of Product Collective, a 15,000+ strong worldwide community of product people. Members collaborate on Slack, meet at INDUSTRY: The Product Conference, listen to Rocketship.fm, learn at Product Lunch and get a weekly brief that includes best practices in product management. In recent years he led business development at DXY, a leading product design firm in the Midwest, and product innovation at MedCity Media, a publishing startup acquired by Breaking Media in 2015.