Innovation in “A Big Company” – Part 3

Are you frustrated that your innovative ideas never seem to go anywhere? Or perhaps you’re an executive wondering how come you are not able to get past the crank-turning and get some innovation into your aging product. Why is successful innovation so hard in large companies? Previously we explained that the initial success of an innovative product leads to an onslaught of demands on resources that make it particularly difficult to continue innovating in the product. Going for too long without some innovation can be catastrophic. We gave some examples of successful evolutionary innovation at EMC, innovations that breathed years of new life into a very mature and successful product. Now we will discuss some ingredients for successful innovation.

Ingredient #1: Understand The Kind of Innovation You Need

Innovation pt3 pic 1Innovation = Idea + Execution. In a clean-slate project (e.g. a startup), the idea part is really important; it has to be big and fundamentally new, revolutionary. The execution part is comparatively easy because it is all new development and there is nothing to slow you down. In the context of an existing product, the degrees of freedom on the ideas are limited because they need to be anchored in the product. So ideas have to be more evolutionary in nature. But they’re there.

Ingredient #2: Identify a Problem and Solution

Three years ago, a product manager and I went on a customer trip to Germany and found something interesting. Every one of the customers asked for a NAS solution with synchronous replication and non-disruptive disaster recovery between two sites. Our existing approach took several minutes to fail over, and that was not sufficient. We saw a clear business opportunity, if we could only deliver it while the competition was distracted.

Now that we had the customer problem, we needed to propose a solution. First, we found that failover took so long because the design was very “physical”. Each Data Mover (NAS gateway) on the source was paired with a passive Data Mover on the target. During a disaster recovery, the passive Data Mover had to boot up before it could take over the load. Second, we brainstormed solutions to this problem. There were the obvious ideas around “can we make the failover process faster” but reducing it from several minutes to a few seconds purely by optimizing the code seemed impossible; we needed a fundamentally different approach. That’s when the product manager came up with the idea. The product had recently introduced Virtual Data Movers which, like a virtual machine on a physical server, abstracted the Data Mover so that it was no longer tied to a physical server. These Virtual Data Movers could be “moved” from one physical Data Mover to another.

Idea: synchronously mirror the Virtual Data Mover and when a failure happens, activate the mirror copy on the other side

Not only did this solve the failover time issue, it also allowed both sites to have active Data Movers versus requiring passive ones, which was another big complaint about our existing solution. A quick check with the architects confirmed that the idea had merit. Next step: execution. This is where most innovation in mature products fails.

Ingredient #3: Show Them The Money

Innovation pt3 pic 2There are many reasons innovation projects fail, but by far the most common one is suffocation. By definition the innovation project is competing with all the other work that is in the backlog for the product to grow, work that has big customer names with $ signs attached to it. Convincing the business to invest in innovation work is challenging. Getting investment is nearly impossible if you cannot assign real and significant revenue against it. In the case of our metro DR innovation, we were in reasonably good shape – we had several big customers with quantifiable revenue and a leading product manager advocating for the feature.

Ingredient #4: PrototypeInnovation pt3 pic 3

Next we had to convince people that the innovation was achievable with reasonable investment – i.e. that the ROI was worth it. Remember, you need to not only show that the project will make money, but that it is worth the opportunity cost (vs. doing something else) and that it won’t become a big sink of engineering resources that puts the core business at risk.

To prove that this was not a high-risk project, we created a prototype. We took a pair of arrays and mirrored them synchronously. An engineer on my AD team Perl scripted the basic logic that accomplished the failover from one side to the other (Note: it helps if you don’t have to go to the engineering team to get the prototype done!) Not only did this validate that there was no technology risk, but it also allowed us to have a very concrete discussion and hand-off with the engineering team that would productize it. The prototype took the senior AD engineer just about a month.

Believe it or not, what we’ve done so far is the easy part, getting the innovation project kicked off – the real challenge is in keeping it going. The longer the project takes to execute, the worse its prospects of surviving and seeing the light of day. In our example, we had projected it would take about a year to roll out – it actually took about two and a half. The main risk comes from the fact that many things can and do change in the organization over such a long time, in particular when you straddle annual planning cycle boundaries. The product strategy can change, macro organizational changes can happen, the people who approved the project may move on, and, of course, the funding profile can (and will) change. These are opportunities for naysayers who didn’t support the idea in the first place to try and squash the project. All of these happened in our case and yet we prevailed. The feature shipped as VDM MetroSync and has logged several large deals and great customer feedback. What’s the key to leaping over all these hurdles?

Ingredient #5: Leadership

Innovation pt3 pic 4Looking back at this and other innovation successes, the common factor I found is that in each case there was a leader, a champion that kept the project alive and pushed it through all the way to the finish line. This is the secret of their success. There are many other factors that contribute too, but without a leader to drive them, I am certain these projects would have failed. In the case of MCx, the project itself was never in real jeopardy of being cut, but without that leader it would have dragged on so long that other macro factors would have eventually suffocated it. The difference the leader made was in sheer execution. They created a can-do culture, demonstrated progress early and often (so that stakeholders can see the big payoff that will come), and galvanized the team to drive harder. In the VDM MetroSync case, execution was not an issue – we had a great team in China that took complete ownership of the feature and showed every sign that they could deliver it successfully. They ran beta programs to ensure the feature would meet customer expectations and based on the feedback, took it upon themselves to build a new UI for the feature that has gotten rave reviews. The challenge was that the project came on the chopping block not once, but several times, as BU strategy shifted and funding cuts happened. Leadership came from the product manager who originally conceived the idea. With some help, he navigated these life-or-death moments in the “board room”. It took a LOT of lobbying, reminding people of the business case behind the feature, highlighting the tremendous progress the development team was making, etc.

Can Do

Starting-from-scratch innovation is comparatively easy – been there, done that. Innovating within the confines of a mature product/business is tricky and very challenging. But it is very rewarding too, in some ways more rewarding. It takes ingenuity of a different kind to pull it off successfully, sort of like those challenges they give kids in the Destination Imagination competitions where you’re handed some newspaper, a few paper clips, some scotch tape and other sundry, and have fifteen minutes to design and build some working contraption with it. But most of all, it takes incredible leadership – unflinching conviction in the idea and superb execution.

-Sudhir Srinivasan @DoctorSudhir

One Reply to “Innovation in “A Big Company” – Part 3”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s