“Failure is the key to success; each mistake teaches us something.” – Morihei Ueshiba
Those Agile methods that you apply to the development cycle don’t just work for the production of code. According to Eric Ries, serial inventor and Entrepreneur-in-Residence at Harvard Business School, you can iterate your way to a business model, learning from your mistakes as you go.
Ries invented the term “Lean Startup” to refer a way of designing a business model through continuous prototyping, rather than meticulous planning. “Using the latest technology, a lean startup can create product prototypes in weeks and months, not years, and use customer feedback to evolve them in near-real time,” Ries said in 2009. “Releases are measured in minutes and hours, not days andweeks.” The intention is to release often, measure everything and learn quickly.
Ries’s own startup, IVMU, releases new production code up to 50 times a day. The code is stability-tested before deployment but there is no in-house user testing. Rather, the development team measure how much the users like the feature by keeping track of the number of shares, uses, abandoned journeys and so on. Features are tweaked, promoted or quietly dropped depending on the results. The great thing about this approach is that you can test multiple versions of a feature on subsets of your users and see which one performs best. The aim is to get the “maximum learning cycles per development dollar” out of the process.
Ries suggests launching with a “minimum viable product” that allows you to recruit your first tranche of users. These users can then be used to test new features while, at the same time, they recruit more users. The model for growth is a gradual mushrooming of features and users in tandem, rather than launching a “complete” product and searching for users to turn into paying customers.
In The Entrepreneur’s Guide to Customer Development, Brant Cooper and Patrick Vlaskovitsexpand on the use of powerful, low-cost and easy-to-use analytics to develop your customer base in the same way that you develop your product: through testing and rapid iterations of marketing and sales strategies. They suggest that customer development should proceed “scientifically” – you should “observe and describe a phenomenon” (in this case, customer behaviour); formulate a hypothesis that predicts customer behaviour; then measure your predictions with formal testing.
Real-time metrics make this experimental approach to customer development a possibility. Again, it is possible to try out several different channels at the same time and measure which ones deliver most value in terms of cost per dollar earned. Dr Florian Heinemann, CEO of Rocket Internet GmbH, is a business analytics expert. He says that the hardest part of building your customer base is understanding which customers will push you to the next level of profitability, and how to reach them. He insists that the gold-standard metric to aim for is the “total lifetime value per referral” for each channel.
“Early in the process you need to have a really good business analytics concept in place to understand who your key customers are…right from the very beginning,” he told HackFwd. It takes three to five months’ worth of data to perform good analytics so the sooner you start to collect data the better.
“In early stages test out every channel possible – even off line – and make it as measurable as possible,” says Florian. “We never do anything that we’re not able to measure.” He even insists on measuring viral activity, pointing out that it doesn’t come for free because someone has to build the viral content.
So how do startups like IMVU and Rocket Internet stay lean and agile?
- They learn from everything: they view each release, and each new marketing operation, as an opportunity to collect metrics and they put smart people to work, figuring out what those metrics tell them.
- They let the customers lead them: Lean proponent and serial entrepreneur Ash Mayura, founder of Spark59, only ever develops features that have “customer pull” – that is, things that his existing users have requested.
- They move quickly: the aim is to maximise learning cycles per dollar, so every time a piece of code is stable enough to be deployed it is put into the field.
- They communicate: they invest in time in social media where they listen to the voices of their customers. It’s less important to broadcast how well you’re doing than it is to listen to how the customers think you’re doing. Take your customer base’s temperature regularly. If they are expressing concerns, engage with them and take them seriously. Your users are your best teachers.
- They apologise gracefully: occasionally you’ll break something when you move this fast. If you spoil a user’s day because you broke their favourite feature, apologise. Explain but don’t make excuses. Show that you have the same aim as they do: you both want you to make a product that they will love using. That customer will be on your side forevermore.
Lean isn’t just for startups: it’s particularly helpful as a growth strategy for organisations that are geared up to be responsive and agile. Even a company the size of Facebook can do Lean, when it comes to product and customer development. As Mark Zuckerberg explained, in a recent letter to Facebook Shareholders: “Moving fast enables us to build more things and learn faster. However, as most companies grow, they slow down too much because they’re more afraid of making mistakes than they are of losing opportunities by moving too slowly. We have a saying: ‘Move fast and break things.’ The idea is that if you never break anything, you’re probably not moving fast enough.”
Remember that not everything you try will be a success. The important thing is to take away a lesson from everything that you do. Keep trying, keep failing, keep learning.