There are a plenty of examples for how to apply Lean startup techniques to starting a new venture or product companies. How about service companies? How can we use Lean startup to grow our business? Here is our experiment using Lean startup for software services company.
About Lean startup
We started by devouring Eric Ries’s, “The Lean startup”, Ash Maurya’s “http://runninglean.co/” books and the video’s at LeanStack. I am not going to repeat all their wisdom. I highly recommend these books and there are a good amount of blogs to get you up to speed with the technique.
Some key differences from a typical product based startup
While for a startup, you are working on getting new customers, for us the company already has customers but wanted to grow by expanding offerings and exploring new customer segments.
Irrespective of the fact if it’s an existing service company or a new one, the typical cycle of delivering software services is quite long. So the challenge here is how do we shorten that feedback loop?
In the beginning, there is the canvas. The recommended time to create the Lean canvas is 20-30mins. It took us 3 hours! But it was worth it. It brought all the people involved on the same page on the most important things for the business – what are our customers problem? How are we solving it today? How can we solve it in the future? What customer segments do we care about? Expected revenue, cost structure etc. It’s highly recommended that this exercise is done with all the stakeholders in the same room.
Grow by how much?
While he dived into the canvas in the order recommended by leanstack.com video’s, we soon realized we were all over the place. The thing that was missing was the goal. So we want to grow, great. By how much? Without this goal, it was difficult to narrow down the options while creating the canvas. E.g.: Do we need to address more problems? How many more customer segments should we go after?
Existing customers and/or new customers?
Without the goal, this was the other problem we debated. Should we have new offerings to our new customers or should we go after new ones? Or both. One could in theory achieve a reasonable revenue goal, by tailoring their offerings to their existing customer base. Of course most of the time, you would want to go after new customers, however, that is not easy nor a cheap proposition. But if you know that is the best way to achieve your goal, then customer acquisition cost is an investment one must make.
Was bit of a struggle. We could go very broad, like Agriculture or get very narrow, agriculturist in my backyard. Finding the right bucket is important. So it could to be a combination of a particular field and scale. E.g.: Agro companies who’s revenue is $200-250 million. Typically a company is already working with customers of a particular scale. That is a good starting point.
While Dave McClure’s pirate metrics (AARRR) provide a good measure of the entire value stream of a startup, a little bit more detail may be required for service based companies acquiring new customers due to the longer feedback cycle. E.g.: We may have to use the customer acquisition lifecycle states (Suspect, lead etc.) while we are working on lead generation. Important thing is, to use just enough detail that is helpful and not add metrics for sake of metrics. The other important thing is to use the classical sales funnel metrics with cohorts for reporting metrics. Here is why.
Revenue and solution offerings
When we got to the revenue section of the canvas, the discussion generated a lot of great solution offering idea’s. In retrospect this seems obvious, as, we all wear our innovator hat when thinking of solutions, however when it does come to revenues that is when we have to make a real connection between solutions and what our customers may find useful and viable. It may be worthwhile to try doing the solution offerings section along with the revenue section of the canvas.
If you are an existing services company (which was in this case), you already have a pretty good idea about the problems your customers are facing. So it’s possible you can conduct problem interviews even before creating your canvas. This can give a real leg up on the proceedings. However, you still would want to spend some time working on the problem statements especially if you are going after new customers as the problems they have may not be same as the ones your existing customer have. A bonus from this exercise is a chance to ask your existing customers, what are the key problems they face. They may surprise us by highlighting things we never thought about.
We have conducted a few rounds of our problem hypothesis interviews and analyzed our validation metrics. It’s already invalidated some of our hypothesis, which if we did not do this exercise we would have used as a basis for creating solution offerings which no one would have bought.
Stay tuned, more to come soon.