Your first idea is always wrong in some meaningful way, argues Scott Anthony. Here’s how to make it right, even if you’re working on a shoestring budget.
Like it or not, you are going to be punched in the face. So how can you take your punches yet keep standing? One tendency can be to turn innovation into an academic exercise. That is, to develop even richer PowerPoint slides, create more refined financial projections, and develop even finer customer segmentation models. Unfortunately, analysis has sharply diminishing returns, especially when you are targeting new or nascent markets that are notoriously difficult to measure.
The best businesses emerge out of trial-and-error experimentation. That doesn’t mean that innovation needs to be random. Instead the best innovators design and execute smart strategic experiments to address their biggest risks. The process is nothing more complicated than what you learned in high school science class: Develop your hypothesis and then go run the experiment.
Here’s a seemingly simple question: How do you know your idea is good? I know how most large companies approach this problem. They break out the spreadsheets. The higher the projected net present value or return on investment, the greater the chance that management will take the idea forward. When you are innovating--particularly if you are doing something that hasn’t been done before--the numbers can be deceiving.
Scott Cook is the founder and chairman of Intuit, whose claim to fame is TurboTax, Quicken, and QuickBooks. Cook is also the source of perhaps my favorite quote about innovation: “For every one of our failures, we had spreadsheets that looked awesome.” Don’t ever confuse an awesome spreadsheet with an awesome business. They are two different things.
So how do you know if an idea is good? Ask five questions:
1. Is it targeting an important problem that customers can’t address because existing solutions are expensive or inconvenient (see my previous Co.Design article for more)?
2. Does it solve the problem in a simpler, more convenient, or more affordable way?
3. Is there a plausible hypothesis about an economically attractive, scalable business model? Don’t believe financial forecasts, but ensure that there’s at least a sensible story.
4. Does the team have the right stuff to course-correct according to in-market learning? Avoid dogmatic teams that will keep trying to prove they are right in spite of mounting evidence to the contrary.
5. Can early profitability be a choice? The sooner there is a line of sight to profits, the better. You might make a strategic decision to be unprofitable by investing in marketing, sales capability, and so on, but at least you know that the core part of the model works.
The primary purpose of this kind of assessment isn’t to generate a score. Rather, it is to get insight into areas where the fit is less obvious or certainty is lower. Those are risk areas that warrant further testing.
What about the numbers? Instead of detailed spreadsheets, reverse the problem. How big does an opportunity have to be to be interesting? What is a simple calculation that crosses that threshold?
For example, a team at a consumer health care company knew any idea had to have the potential to cross $100 million in annual gross revenue to get support from leadership. The population of severe sufferers for this particular condition was about 10 million people. The product the team was thinking about introducing would cost $20 per package. The team’s best guess was that the average consumer would purchase five packages a year. So that’s $1 billion if the team penetrated the entire market. That meant that getting 1 million people--10% of the market--would allow the team to cross the magic $100 million mark.
This is the innovator’s version of marketing’s famous four Ps (price, product, place, and promotion). Getting to $100 million in revenue required penetrating 10% of a population of 10 million who purchased the product five times a year at a price of $20 each.
Two pieces of advice: First, be precise about your target population. Don’t fall into the trap of defining the market so broadly that any threshold is achievable (“If we just got $1 from everyone in India …”). Create as small a market as possible, notably, the people who would constitute your dream customers. Second, don’t assert a penetration rate--solve for it after you’ve estimated the population, pricing, and purchase frequency.
This deceptively simple calculation neatly captures many of the elements of an idea’s business model. Does the idea target a niche or a mass population? Is it an occasional or frequent purchase? It also surfaces key operational assumptions. What channel would support the target price point? What kind of post-sales service would be necessary, given the purchase frequency?
These exercises help you zero in on the most critical assumptions behind success. Your hypothesis of course is that these assumptions are true. That’s just a hypothesis until you design and execute your experiments.
The plan looked great on paper (plans have a way of doing this). Our team at Innosight was going to create a new business to attack the “missing middle” in the Indian men’s grooming market. A single visit to India showed me the promise of this opportunity. If you wanted a shave or haircut, you could go to a high-end salon at a six-star hotel and get a truly world-class experience. Or, you could get incredibly affordable service from the barber whose “salon” was a single chair that sat alongside the road. His instruments at least looked not-too-unhygienic. But if you wanted something in between--a solid experience at a reasonable price--you were pretty much out of luck.
We called our idea Razor Rave. The plan involved an innovative retail store format that essentially put a single barber chair inside a small pod. The pod’s small footprint provided low overheads and high degrees of flexibility. We would use world-class products and envisioned tie-ups with the Gillettes and L’Oreals of the world.
So would consumers be interested in Razor Rave? There was only one way to find out. We rented a truck and created a salon on wheels by putting a barber’s chair on the back of the truck. We drove the truck around the streets of Bangalore for a couple of weeks. High levels of consumer interest told us that the market was ready to pay price premiums over the roadside barber or local low-end salons. Total cost of learning? About $3,000. We were off!
This is an example of a targeted experiment, where you isolate a specific variable and test it (see Matt Eyring and Clark Gilbert, “Beating the Odds When You Launch a New Venture,” for more). Targeted experiments work well when there is a clear, identified risk that can be tested directly.
When you are doing something new, however, some of the most critical assumptions can’t be isolated. You can only learn about them by putting the entire business together. You have to run an integrated experiment, such as a pilot or a simulation. An integrated experiment doesn’t have to produce huge results. The point is to learn about the “unknown unknowns”--or the things you didn’t know you didn’t know.
Our integrated experiment for Razor Rave highlighted what we ended up dubbing the “hero barber” problem. By launching a few pods on the streets of Bangalore, we quickly learned that having a good barber was absolutely critical to drawing enough customers to make the business interesting. A good barber already had a loyal group of customers who would go out of their way to frequent the Razor Rave pod. These barbers were able to take new customers who wandered in, intrigued by the pod’s format, and turn them into repeat customers.
The hair cutters were, in fact, too good. You see, the single-chair format clearly made the barber a hero. Once the barber figured out how critical he was to the success of the business model, the demands started coming. We were left with the unappealing choice of increasing the barber’s wages to the point where our economic model began to fall apart, or suffering high attrition. We couldn’t find an obvious way out of this quandary without completely redoing the business model, so we decided that it was closing time at the Rave.
One final tip--it’s easy when you are testing an idea to fall into a trap psychologists call confirmation bias. The singer Paul Simon provided a simple definition of confirmation bias in his song "The Boxer": “Still a man hears what he wants to hear and disregards the rest.” It’s easy to focus on the things you expect and ignore the things that are surprises. But remember--your first plan is wrong. The most interesting findings are the ones you didn’t expect. You aren’t trying to confirm things you already know, you are trying to discover things you didn’t expect.
The single best way to spot surprises is to personally participate in experiments. If you delegate the task, you get a nice glossy report that answers your question, but misses the point. Personal participation with an open mind helps you spot surprises that point the way to long-term success.
***
There’s much more to say about the topic of testing, and the writings of Rita McGrath, Peter Sims, Steve Blank, and Eric Ries are replete with practical tips. The most important thing is to just do it. Don’t be an Ivory Tower Innovator who keeps trying to sell the world on a 60-page PowerPoint document with neatly arranged 10-point bullet points. Whenever you have an idea, look for the quickest way to learn about critical assumptions--or to find out the critical assumption you didn’t realize you were making. Remember Thomas Edison’s guidance: “Genius is 1% inspiration and 99% perspiration.” If you aren’t sweating, you aren’t innovating. Get to it!
No comments:
Post a Comment