In a 2018 Dreamforce workshop called “Experiment Like a Startup to Unlock New Value Faster," participants were led through an exercise by workshop leaders Sherri Hutter and Kyleigh Wawak of Salesforce that would show them how they could use small experiments to help unlock business value.
Most people think about innovation as a sort of funnel, said Hutter and Wawak—a bunch of ideas that gradually get winnowed down to a final candidate. That model works for incremental changes, they said, but true innovation is more like a maze with multiple choices and possible directions. Innovation tends to be hampered by the four Ps, they said: PowerPoint, Persuasion, Politicking and HIPPO (highest paid person's opinion). Because of those factors, most innovative ideas never see the light of day—and even when they do, they're probably not in their best form right from the start.
As an example, the presenters offered the story of a car sharing service. Rather than doing focus groups or surveys, the startup's founders went right to testing their hypothesis that the idea would work: They put cars on the street to see how people responded. Businesses could apply that lesson, the presenters said, by designing micropilots: fast, lightweight experiments to validate a hypothesis with real-world testing. In an iterative process, a micropilot involves building as small an experiment as needed to test—creating the “minimum believable prototype." Whereas a full-scale pilot is a live test of a complete product or service, micropilots target specific key questions.
Feasibility and viability only matter if people want your offering.
—Sherri Hutter and Kyleigh Wawak, Salesforce
Designing a micropilot involves identifying uncertainties in your business processes or products, developing a hypothesis around it, and determining how you would measure the result of the pilot. The presenters described five possible ways of testing an idea:
- Manually simulating a new offering without automating it yet.
- An explainer video, which outlines user benefits and invites viewers to sign up for the beta launch.
- A pop-up shop, or a branded kiosk where people can try a product or service and give feedback.
- Crowdfunding, or asking people to put up money in advance to demonstrate demand.
- A “false door," i.e. a branded web page that describes a product that doesn't really exist yet.
With that information in hand, workshop participants moved on to practice designing a micropilot. The presenters distributed an “innovation brief" describing the idea for a wine and spirits subscription program that would appeal to millennials along with some mock feedback from potential customers. Participants were asked to come up with a hypothesis and an experiment to test customer concerns as expressed in the feedback. For example, whether someone would really stick with a subscription over time. The presenters advised that the experiment should be one that could be carried out in a day or a week and that would contribute to an iterative process. They also told participants not to be worried about the experiment failing—“test to learn and shape, not to justify," they said. “Call it 'learning,' not 'failure.'"
After the participants shared the results of their exercises, Hutter and Wawak offered five tips for running experiments on your own:
- From the Venn diagram of the feasibility, viability and desirability aspects of an idea, always start with desirability. “Feasibility and viability only matter if people want your offering," they said.
- Have focused learning goals and test a specific question.
- Keep prototypes lightweight—build the “minimum believable prototype."
- Aim to learn something surprising, and be prepared to fail. Translate what you learn to refine the idea.
Photo: Getty Images