Tullman: How to Innovate Like a Startup

By 1871 CEO Howard A. Tullman. To view the original blog, visit: http://www.inc.com/howard-tullman/sneakernet-innovation-technology.html

Innovation at companies of all sizes requires a willingness to test out inelegant solutions. In other words, break out your sneakers.

Many years ago, I saw a great cartoon that completely encapsulated the leap of faith that is a crucial part of every entrepreneur’s dream. The drawing showed a guy standing under a series of thought bubbles reflecting the development of his great new business idea, from creation through development and on to commercialization. Right smack in the middle of the process was a modest little bubble, which read “miracle happens here.” For me, that simple illustration said everything there was to say about how critical faith and perseverance are to the success of a startup. And how sometimes, it’s not about seeing; it’s about believing. You have to believe it and concretely envision it, then, you can convince others to see it as well.

In addition to that initial takeaway, I have retained–all these years later–a completely different thought about that critical link in the process. This has helped me to understand and formulate inexpensive, practical, and rapid solutions for small and large businesses that can help move things forward without asking anyone to bet the farm. Mine isn’t a new concept, in some respects; we’ve all heard of “try it before you buy it.” And I used to remind my MBA students that you never wanted to test the depth of a puddle by jumping in with both feet. Making cheap mistakes–and getting over them–is an art form.

But this simple concept continues to be a major challenge for large businesses that want to introduce new ideas and innovations but are also grappling with the need to deploy enterprise-wide solutions and manage the tangled code of decades-old legacy systems. They long for the speed, flexibility, and low costs of the startup world, but they’re stuck in an environment where the passage of time and the prior investment of millions of dollars makes it harder and harder to change. They need an approach that will let them investigate and iterate before they invest and adopt new solutions. And they need a buffer to be sure that as they’re testing new tools they aren’t disrupting the ongoing business.

My solution–which I call a “sneakernet“–isn’t groundbreaking or revolutionary–it’s just the heart and soul of how startups think about these things. Large businesses need to think like this to get over that first hurdle of doing something “good enough:” quick and dirty–with duct tape galore–in order to get things going. It won’t be pretty–it certainly won’t be perfect–but if you wait until it is, you’ll never even get started.

As with everything, it starts with a definition. I’ve found that properly defining the problem often gets you more than halfway to the solution. And, as often as not, the solution that you need to get the ball rolling is something that’s painfully obvious and sitting right in front of you. Very often, it’s not a step or two forward; it’s a look backwards at how things used to work and how they were done before all those expensive computers got in the way.

 So what exactly is a sneakernet? 

It’s a back-to-basics perspective that recognizes that sometimes shoe leather and sneakers are better system connectors than six months of re-engineering, thousands of wasted man hours, and unending attempts to get two separate systems to talk effectively to each other. It’s when you can’t solve things the hard way, so you take a much shorter and simpler path and just walk the data from one system over to the next. And even though you may have to re-enter critical information and there may be some redundancies, you’ll get the job done, as far as these tests are concerned, and you’ll actually save time, money, and a lot of sleepless and sweaty nights.

This is an approach that doesn’t try to solve 100 percent of the problem on Day One. It doesn’t attempt to create a comprehensive and complete process (which will cost too much, will take too long, and will most likely never see the light of day). It’s an approach that says, if I was starting from scratch and just had to get the thing from here to there, how would I do it?

If you can’t breach the four walls of an enterprise or legacy system and get your job done directly, the next best attack is to step back and do things the old-fashioned way for a while, especially when you’re just trying something new. For example, if you can’t readily or inexpensively integrate a new mobile ordering system with your old payment system, then do it the easy way. Capture the orders quickly and easily from the new mobile system (thereby making your customers a lot happier), print them out, and then quickly walk them over and re-enter them into the old payment system, which will take it from there. That’s a classic sneakernet approach.

Sure, it won’t scale, but who really cares? For the moment, you’re just trying to see if the dogs like the new dog food. No one really needs to know what’s behind the curtain. It’s really not important how the critical information moves or who moves it; it’s just critical that the information move.

• This approach is smart, because you limit your investment and your dedicated resources while you are still able to effectively test the consumer’s interest and appetite in the new approach.
• It’s secure, because there are none of the data issues that a real-time connection might involve–and which would give your IT people ulcers for sure.
• It’s swift, because it is supplementary to everything you are now doing and can be immediately implemented without any material changes or integration into your current operations. To those ancient and immovable legacy systems, it looks exactly like the “same old, same old,” but you’re actually appending new solutions to the system.

It’s a sneaky sneakernet.