Thursday, 21 September, 2017
If you try to grow your online business based on gut decision your either wrong or just plain lucky?! If you on the other hand want to grow your business today, tomorrow and the day after tomorrow I recommend you to go the scientific way!?
This Tuesday we had the great pleasure have a real startup super hero visiting us, Gustaf Alströmer Y Combinator partner and former head of growth at Airbnb, talking about growth. There were plenty of curious entrepreneurs like myself wanting to know more about this topic on how to grow your online company. Here comes some of the things I picked up.
Growth is very situational and dependent on your current context. But it all starts with having a product that customers really love! Without having a product that customers actually are using you’re only trying to fuel a fire that won’t start.
Start with getting your product market fit right! Your goal is to make something that is useful and delights your customers. To reach this you must measure things and even more importantly measure the correct thing, I repeat thing not things! So find/define that specific metric that represents the value. For example, AirBnb uses “bookings” whereas E-Bay uses “gross value of sales” once that is in order you should decide on what frequency to combine with, e.g. per day, per week, per month etc. This metric of choice may sometimes change over time as your business grows.
One rookie mistake is that you start measuring new users for example, so if you have new users signing up everything is kicking?! New users signing up says basically nothing about your service or if your customers like it nor not?! What you need to measure is retention, how many are using your service and returning to use it? This could be called growth accounting just like traditional financial accounting but for growth instead and it is the future of growth.
When you have metric of choice in order and your growth accounting up and running is when you can start to do experiments. Experimentation is all about coming up with new ways to improve your metric of choice and of course you typically don’t change everything at once… There’s no use in setting up experiments if you don’t also set up a counter factual, this is basically A/B testing. You let some users either see an unchanged version of your webpage/e-mail etc. or your changed version and then compare the statistics for these.
How many experiments you set up and run and how big they are differs from company to company, but Candy Crush Saga usually runs at least 300 parallel experiments….. 300 parallel experiments might be to many for you at this point but doing three experiments a year really doesn’t cut it. One out of five experiments is a success, one out of five is a slight success and the rest, three out of five is crap, so in order to learn and to grow you need to do a lot of experiments.
Want to know more on how to start your startup or grow your existing venture contact me or any of my colleagues at Chalmers Ventures and we will be glad to be at your service.