The concierge mvp is an interesting concept because the lean startup method is all about eliminating waste and the concierge mvp seems wasteful from a distance. To a software developer, it doesn’t make sense to manually perform a task when you can get a software program to do it for you.
In one of Eric Ries’ posts, he shows how small batches can actually help you become more efficient. In the post, Eric tells the story of a father who has a content with this children to find the fastest way to mail a letter. The children suggested folding all the letters first, then stuffing the envelopes first, then sealing the envelopes, and finally inserting a stamp. The father suggested to mail each letter one at a time. Ultimately, the father wins. He wins because of the time it takes for the children to sort and stack after each step. There is more explanation in Eric’s post.
Real World Implications of Small Batches
Besides a simple contest between a father and children, there are many real world implications for small batches. When developing a business hypotheses, the successive steps are not always known. It’s simple to fold a letter, but it’s difficult to build a software project. For the first version of a software products, developers aren’t always sure of which feature should go next in a linear process. They often use their experiences to guess what a user wants next. But this is usually wrong. Often, a step in a software product will be the equivalent of stuffing a cucumber into the envelope instead of a letter. In hindsight, some software decisions are obviously wrong, but they don’t always have to be. In order to understand how wrong certain products can be, you need to walk with your customer through the problem/solution stages. There is no better way of doing this than using the concierge minimum viable product technique.
The Concierge Minimum Viable Product
If you’re unfamiliar with the concierge minimum product, your world is about to be turned upside down. The concierge mvp is a minimum viable product where you manually guide your user through the solution to a problem.
An example would be an automatic couponing program. Let’s say you want a software program that will automatically send a user coupons based on the food they buy each week and help them decide which grocery store they will shop at to save more money. Instead of building software, first you would allow a user to tell you what they buy each week, maybe through an email or face to face, and then take the coupons and best grocery store to them each week. Soon, you’ll find out if/when the user doesn’t go to the grocery store, if savings really affects which grocery store they choose, and if they care about certain brands, and if so, in which food categories. You would learn a lot more by using this concierge mvp technique than by taking the enormous effort of building the web application. This will allow you to decide which ideas work from your initial hypothesis, and which ideas need to be scrapped.
The concierge minimum viable product is inefficient at solving a problem, but it’s not a long term solution for customers. It’s a short term solution to help you learn how to solve customer’s problems. The point of the concierge mvp, as the point with most mvp’s, is to maximize learning and mitigate risk of developing a crappy product. The point is to use the concierge mvp so that you don’t try to stuff a cucumber into an envelope.
Food on the Table is the current known best example of following the concierge mvp technique. Here’s a video of food on the table’s experiences.