Testing business ideas

At Imperial, we are familiar with scientific experiments taking place in the labs. This is essential to make sure medicines work without adverse side effects, materials are strong enough for construction and new fertilisers do more good than harm.

This evidence-based harder, more time- and cashapproach is also consuming experiments with essential to building stronger evidence. a startup. Failing to test your David learnt the business idea could result importance of testing in a lot of time and money business ideas in his early wasted. We caught up with career, which consisted David J Bland, co-author of of working with various Testing Business Ideas, to startups, some more find out about the best ways successful than others. One to test, deciding what counts was acquired for $16 million, as ‘success’ and knowing which, as David describes, when to quit ... was “a pretty decent exit for

We’ve heard before 2006”. But following this that failure is a big part of success, David joined a few entrepreneurship. And whilst startups that didn’t go quite many founders can tell you so well. “What I learned was a story or two of when it that if you blindly ignore the all goes wrong, there are data and keep persevering ways to limit risk. You may no matter what, you have heard the term ‘failing eventually run out of money. fast’: testing, learning from Basically, it doesn’t matter the results and changing how amazing you think it is, direction when you know it’s if your customers don’t want not working, before you’ve it, they don’t want it.” put too much time or money This realisation sparked behind an idea. David David’s journey to help Bland, serial entrepreneur other founders. David and Silicon Valley stalwart, went on to work with has made it his mission accelerators and large global to use his experience to companies to encourage an teach fellow founders about environment of innovation the importance of testing and experimentation where business ideas before you he crossed paths with get in too deep. His latest Alexander Osterwalder, book, Testing Business creator of the Business Ideas, co-written with Swiss Model Canvas. “All these business model strategist founders or small innovation Alexander Osterwalder, is an teams just weren’t aware of experiment library consisting all the different experiments of over 40 illustrated that are available. So, we business tests, organised ended up writing a book. We by cost, time and strength wanted to help people say, of evidence, that guide ‘I’m stuck with something. founders as they build an What can I do to help me evidence-backed business make progress?’” ready to scale. David recommends a founder Getting started start with quick, easy For founders, the first step to experiments that are free or testing a business idea lies cheap, and elicit relatively in the value proposition and weak evidence, and then, customer segments. David if the evidence validates explains: “You need to look their hypothesis, move onto for these unmet needs or a

Don’t get too excited if one experiment succeeds, and don’t get too depressed if it fails.

way to create something of value that’s different.” One of the most common mistakes entrepreneurs make is only making moderate improvements on what’s already out there. Customers are unlikely to change habits for a slight improvement. “It has to be 10 times better,” David says, “and don’t build right away. There’s so much you can learn without building anything.”

One of the biggest pitfalls for startups testing their idea is bias, David explains. “Usually all these biases creep in, like confirmation bias and experimenter’s bias and overconfidence bias. And what happens is, you will run the experiment, but then you only look for the data that proves what you want it to already do.” David says the one thing he wishes he’d known starting out was: “to have strong opinions held loosely. So this idea of having an opinion but be open to the idea of being wrong and adjusting accordingly.”

Once you have your value proposition and customer segment defined, David explains that the next step is assumptions mapping: “The flow I typically follow is to sketch out your strategy. Write down your assumptions map and then run experiments on the really big things that if proven wrong, would cause everything else to fail.” Assumptions are often where the biggest risks lie, so finding tests that can prove or disprove them is vital. “There’s all kinds of ways we can get in front of customers. It could be interviews or landing pages or surveys, but there’s so much more than that. You can do paper prototyping or a clickable mock-up or even 3D printing. There are all these different tools you can use to learn and generate evidence that don’t require building hardware or software.”

The varying experiments in David’s book not only allow entrepreneurs to test the viability of their idea, but also the feasibility of running their business model. An example David gives is: “There’s always machine learning startups that struggle to hire data scientists. That’s a big risk that’s on the back end of your startup if you’re trying to do machine learning and you don’t have any data scientists. Your risk isn’t the customer there. It’s the backstage, the functioning. You also need to test onboarding people. How do you attract talent? How do you run the infrastructure? Key partners? Or test out pricing? It’s more than just customer. Customer is important, but there’s much more to it than that to be successful.”

Next steps

After testing all these aspects of your business, a decision needs to be made as to whether you move forward, iterate or pivot, based on the success of the tests. David advises writing down what you’re going to measure and the success criteria prior to conducting the test. Always be ambitious, he counsels. “If we’re working on something new, you don’t want to barely succeed, you want to challenge yourselves, so I tend to set the criteria somewhat high.”

Even with numerous successful business tests, taking the leap can be daunting. David says it’s important to remember: “Don’t get too excited if one experiment succeeds, and don’t get too depressed if it fails. Usually I run various tests for about 12 weeks at least before making that decision. Twelve weeks is kind of arbitrary, I admit. But I feel like that’s enough, that if you’ve tested one thing every week for 12 weeks and there’s just nothing there then you should probably rethink your strategy. If you’ve tested for 12 weeks and you’re onto something and have traction, then that could help you make that decision to take the leap.”

Despite the obvious benefits to continued testing, big corporates are unlikely to have a culture of experimentation if no longer run by the founders: “They get in this mindset of just execute business as usual.” Despite this, a few large companies have found ways to test new products without risking their brands’ reputation: “You have to think about how to create the ‘new thing’, so the teams I work with and I’ve observed, will do something like lightly branded, or off brand, or a labs brand. If you go to Indiegogo and scroll through it, you’ll see a lot of corporate ideas that are on Indiegogo right now, it’s pretty interesting.”

One example of this is a well-known software company in San Francisco, USA, which David has worked with to launch new products: “I’ve helped launch some of their new mobile apps, and I felt like they did a great job of throwing a team together that’s dedicated at something, testing it off brand, seeing if it works, and then bringing it on brand, which I thought was very smart. If they launched something with their brand that’s not polished, they’ll get a lot of press really quickly on it. There’s all kinds of cool stuff happening, it really comes

Three types of testing: 1 Wizard of Oz

A test where it appears to work seamlessly on the front end, but manual work happens behind the scenes. For example, a landing page. 2

Concierge

All done manually, like a white glove service, delivering the value to customers. 3

Mash-up

A combination of the two. Cobbling together a solution that still adds value. because it’ll take three to five years to really help it start to stick and in the startup that’s like dog years.”

For startups and large companies alike, one concern when testing is protecting your idea and IP. David agrees that this can be a risk: “It’s a challenge. Sometimes we have people sign disclosures or an NDA to test with us. They’re all these sites where people want to launch and test ideas like Product Hunt. But if you launch something on there that you’re not quite ready to run with there are a lot of other hungry entrepreneurs scrolling that list. So you do have to balance it. But there’s a lot you can test without exposing all your IP.”

The most successful businesses have always been based on evidence. Evidence that their customers want the product or service, evidence your company infrastructure will work and evidence of your value proposition. So, run experiments. Use experience in research to validate your business idea, and remember: “be open to the idea of being wrong and adjusting accordingly”.

David’s book, Testing Business Ideas is available from the Enterprise Lab Library and on Amazon.