Probably more like 5 years. AI has arrived (for real this time), and it promises to automate away mundane work, make great use of your big data, make humans more productive, and create efficiencies and value at a scale the world has never seen before. That’s a lot of corporate evolution. And unfortunately the way the universe deals best with evolution is through good old natural selection. The old and unadapted die, and the new carry forward what’s left. Enterprises just don’t evolve quickly…
John Chambers, the CEO of Cisco, told attendees at their annual conference that 40% of them will not exist in a meaningful way in 10 years. By the way, I wouldn’t exclude Cisco from that list.
If you are one of the lucky evolution survivors, you are not out of the woods yet. There’s a never-ending supply of predators headed your way since AI and modern technology will continue to lower the barrier to entry for competing products and ventures. Meanwhile, the rate of change of your customer needs, market opportunities, and technology capabilities will grow exponentially. If you don’t keep up, you won’t be around for long to worry about it. In order to continue to survive and win, you need to offer increasingly differentiated customer experiences, and quickly.
Off-the-shelf SaaS is in trouble. For most customer-facing software, it is (or will soon be) cheaper, faster, and better to build custom software than buy and customize SaaS products. This has been a long time coming thanks to open source, cloud, mobile, big data, agile, etc., and AI is the next nail in the coffin.
Having the same SaaS software as everyone else not only leaves you stuck in the middle of the pack without the ability to differentiate (see #2), it’s also pretty dangerous. That SaaS software you depend on today to run critical aspects of your business and to interface with your customers, vendors, etc., could literally disappear overnight (see #1). Over the last decade, we’ve seen companies like Parse, Homejoy, PayByTouch, Omnidrive, Fab, Ning, Joost, and SearchMe leave their customers hanging, and I believe this is just the beginning. At least a few of the nice ones, like Parse, left customers with a somewhat viable open source alternative when they went away, but that’s only because Facebook gave them a soft landing. Don’t expect the others to be that nice.
According to Philippe Botteri at Accel, “It takes around 10 years for a SaaS company to reach maturity”. In an AI-native world, 10 years is an eternity.
Most companies will not control the applications that their customers use to interact with them in the future. The digital and real-world environments and devices where users spend most of their time will own this. In the digital world that’s Facebook, WeChat, Google, Amazon, Apple, Slack, etc. In the real world, it’s devices like Echo, Nest, TV’s, VR/AR, robots, etc. If you are not building your software to interact with your customers seamlessly across all of these channels, one of your competitors is. Your customers want you to meet them where they already are, not the other way around.
In 10 years we will joke, “Remember when companies used to tell us to go to their website and download their app? Haha!”
Yes, pretty much all of it. This is not just about duct taping a few more modules and some IBM Watson calls onto your existing enterprise software. Most of the world’s software actually needs to be re-architected and rewritten in order to leverage AI and become omni-channel in a way that can evolve quickly enough. If you are not building on a polyglot micro-service architecture that supports domain-specific inter-service and inter-device messaging, containerization, embedded analytics, automated metrics-driven development, testing, deployment, and cloud operations, you have some work to do!
AI-native software should also be built on open platforms and be portable across clouds in order to avoid the vendor dependency risk (see #3). Whether you are designing, building, and running your software yourself (which I will argue in #7 you probably can’t), you should at least have the ability to at any time. “Nobody ever got fired for buying IBM” is about to change in a big way. If you don’t have the source code to all of the software your company is dependent on, your company might not be around long enough to fire you.
Most people don’t think about the fact that the majority of customer facing software is pretty much the same across companies, and even industries. Developers spend so much of their time building and debugging the same stuff as each other: user account management, payment processing, content management, caching, feeds, push notifications, and so on. Fortunately more and more of these building blocks are available as open source. As long as you assemble them intelligently, the community can write and maintain a lot of great code for you for free.
This leaves the special parts of your software: your user experience, branding, domain model, business logic, workflows, secret sauce, killer features, core intelligence, etc. This is generally less than 20% of the code. AI is going to take this down to less than 5%. Machine learning allows your software to become smarter on its own by enhancing its own code. Domain models will define themselves, business logic will figure itself out, and the UX and branding might even begin to evolve itself based on how your customers interact. Watch for a lot more reusability of software across verticals (“horizontal is the new vertical?”).
So what is left? The last 5% is the most creative, most differentiating, and most unique parts of your software. The good news is that you can now redirect the budget you are spending on the rest of your software to just this 5%. But since that’s all you have to beat your competition, you better make sure your software platform is incredibly flexible to enable you to focus on doing anything you want with that 5%, and that your software talent is great enough to actually do it.
Designing, building, operating, and evolving the type of AI-native software I describe above is really hard. It requires the right technology and platform, good best practices, a clear business and product strategy, good leadership, and most importantly, access to great talent.
In software, great talent is at least 10x more productive than good talent (and infinitely more productive than bad talent). Everyone seems to believe this, but I rarely see companies live by it. I’ve literally had very successful CEOs tell me that they realize a team is producing 3x more than others, but they cost twice as much so they need to replace them with cheaper talent. Not only is that bad math, but talent is not talent. It’s not about what people know or what’s on their resume. It’s about their intelligence, potential, drive, speed, diversity, lateral thinking, problem solving, ability to take risks, capacity to learn, and overall mental bandwidth. Unfortunately, there’s just not a lot of this type of extraordinary talent in the world.
If you are going to overcome challenges #1–6 and break out of the pack, you will need this 10x talent. Unless you can provide a Silicon Valley culture of constant learning and challenges, unlimited flexibility, open innovation, attractive geography, extraordinary mentors, an entrepreneurial career path, and a higher purpose, you should feel lucky to have a couple employees that fit this bill. Note that I didn’t list “exorbitant salary”. While you do need to pay people fairly, throwing ridiculous money at extraordinary talent rarely works out. You might get a year or two of good (not great) work out of them, but without the right culture, you are not going to get 10x and they won’t stick around for long. This talent-first culture is vastly different than most enterprise cultures today.
When I speak to most enterprise CEOs about this, their reaction is usually something like “I already know all of this, but it’s not going to happen tomorrow and we are too focused on big initiatives right now to worry about it.” Plus they have an (underfunded, non-empowered) innovation lab/division thinking about it. Needless to say, there’s no sense of urgency.
When I speak to non-enterprise people, their reaction is overwhelmingly “These old companies all just need to die to make room for the next generation. They can’t be fixed. You saw what happened to taxi companies.” The impending doom couldn’t be more clear to them.
Unfortunately I don’t have all the answers and this second group may well be right. But I do believe there could be something in the middle. Enterprises have so many amazing assets in their customers, partnerships, people, revenue, profit, products, experience, domain knowledge, sales channels, relationships, goodwill, etc. I think there is a way to break a company down and build it back up as an AI-native company. But I will leave that idea for a future article…
Hacker Noon is how hackers start their afternoons. We’re a part of the @AMIfamily. We are now accepting submissions and happy to discuss advertising &sponsorship opportunities.
To learn more, read our about page, like/message us on Facebook, or simply, tweet/DM @HackerNoon.
If you enjoyed this story, we recommend reading our latest tech stories and trending tech stories. Until next time, don’t take the realities of the world for granted!