Companies across industries are vying for a competitive edge by getting innovative products into the hands of their customers—fast. However, when designing and piloting new digital products in the market, our clients in large, established organizations often face competition from lean startups that aren’t encumbered with legacy IT infrastructure. Entrenched software platforms with confusing acronyms are inevitably blamed as the limiters—the bottlenecks, the crushers of dreams. They shouldn’t be. Want to compete with the startups? Start building products like one.
A pilot is a way to test new or disruptive ideas in the market—ideally as quickly as possible. If the pilot is successful, then perhaps IT will need to solution some legacy integrations. Or not. Maybe the systems in question will be deprecated by the time the pilot ends. Maybe analytics captured from the pilot will cause a pivot in the experience design, changing the required integrations. The point is: Thinking too far ahead when piloting a product can get you stuck in the mud.
Our clients often dread the software development process, scarred by past large-scale integrations and cross-department rigor mortis. Don’t lose sight that a successful pilot is experimental and should be an auxiliary to your established operations and technology. If a planning conversation turns to, “But how will this product leverage [insert legacy system],” the point may be valid for widescale deployment, but probably not relevant to a pilot.
During pilot planning, our clients are often tempted to lean on existing platforms and vendors, thinking that it’ll be easier, more cost effective, or will save time. Depending on the requirements, failing to look outside the familiar toolbox could result in nailing a square peg in round hole, or killing an ant with a sledgehammer.
Let’s face it, most software applications contend with the same set of functional hurdles: authentication, user communication (email, push, SMS) and payments. Technology teams also face a familiar set of tasks: server provisioning, automated testing, deployment, and monitoring.
The good news is that over the past decade an entire software industry has blossomed to solve the common functional needs of every type of digital product. These “cloud software” companies are really in the business of selling “the wheel” so that you don’t have to reinvent it for every project. Your product needs authentication? Auth0 or Okta. Send automated emails? SendGrid. Send SMS or push notifications? Twilio. Payments or subscriptions? Stripe or Braintree.
Application frameworks aimed at startups—namely, Amazon’s Amplify or Google’s Firebase—can accelerate development and DevOps tasks (i.e., getting to production). Both offer simple solutions for essential services like storage and analytics, plus a fast lane to deployment on AWS and Google Cloud, respectively.
With a smart approach to leveraging the vast ocean of cloud services to accomplish the fundamentals, developing custom software with venture speed is within reach. Define the requirements, and then look at all the functionality you can offload to a cloud service. You may find that 25 percent or 50 percent of the application can be developed following an implementation guide (just hooking the right wires together), leaving a much smaller core of custom development.
Many cloud service offerings, including those from AWS, Twilio, or Stripe, use the popular pay-as-you-go model, which is perfect for pilots, where service utilization may be small or unknown. It’s little up-front investment—and that’s the point. The competition among these cloud service companies is fierce, and they want to provide every incentive for customers to try out their products. They’ve lowered the barrier to entry, creating a win-win for these companies and their clients.
Disruptive SaaS companies such as Stripe and Auth0 don’t sell their products with a huge marketing budget or over a round of golf with the CIO. They sell to developers (the folks that use the product) with no upfront cost, simple APIs, and awesome documentation. The hope is that their popularity will grow with the hobbyist and hacker crowd and then bubble up to the enterprises that employ them. So far the strategy has been successful (10 years in, Stripe services more than one million businesses). The point is, your developers probably know and use these platforms, further reducing the implementation hurdles.
We know from experience with our own venture programs and pilots that many common software routines, like authenticating or sending SMS can be accomplished with a few dozen lines of code. So don’t be afraid of custom software. Much of the hard stuff has already been figured out. The real hurdle is designing a digital product that people will use—that’s sticky—and will ultimately provide measurable positive impact on the organization’s goals.
Remember, the goal in testing these innovative new products for impact is to see how (and if) people use the product. If the pilot is successful, then push for the long-term investments that pay off big. And if the pilot doesn’t gain traction? You’ve learned a lot with a little investment. The next pilot will be easier. Oh, and you don’t have to figure out how to undo that legacy integration.