Jan Dawson wrote an interesting analysis of Android version adoption trends that’s worth reading. There’s a small comparison to iOS I’d like you to pay attention to. In short, Android updates top out at about 40% of the base somewhere about 18 months after release — six months after the next release is available. iOS updates hit about 90% adoption within the first year. Go ahead, read it, then come back.
Something struck me as I read this in relation to enterprise software. Took me a while to put my finger on it. It’s important though because I think it gives us insight into the sea-change that’s happening in technology; the change that I describe as a problem of scale.
The pace of change vs our ability to absorb new technology
At an earlier middleware company in my career, I talked to customers about their ability to absorb new releases. Generally speaking, they could absorb a new release every year or so. That’s counting from the time of first starting to learn about the new release and evaluate it, to negotiating terms, to new development, testing, and release scheduling.
It’s a long process, but one that informed our own release schedule objectives. If we had a new release as “quickly” as once a year, the teams responsible for our software would spend their whole lives managing upgrades. Talk about an expensive investment just to keep the lights on.
Not every piece of software is as complicated or has as many dependencies as middleware. So perhaps middleware is the exception, not the rule. But this:
Now look at the difference in adoption of new mobile OSes highlighted in Jan’s article. The latest Android release is expected to peak at about 40% penetration, some six months AFTER the next release. Compare that to iOS which sees some 90%+ adoption within the first year.
It’s no wonder that iOS users (more importantly, their expectations and behaviors) are stressing IT organizations. So much of the base has access to new features and they want the software they use to take advantage of them. From screen size optimization, to touchID, to sharesheets, and so on. These are the same IT organizations that think it’s OK to still run Windows XP in places.
But IT isn’t used to moving that quickly. And in fact, in many cases can’t without significant organizational and cultural change.
The future is going to require even more scale
Does this mean mobile will “break” IT?
What about IoT?
Whatever IoT means, it’s going to mean even more software than mobile because like mobile:PC, IoT:mobile represents an order of magnitude change in the number of computing devices IT has to accommodate. As a result of that order of magnitude change, there will be a lot more software experiences required. Even more software than required by the shift to mobile, which I believe IT isn’t managing so well.
An alternate point of view
Of course, there’s an alternate stat that’s worth considering:
Maybe I’m wrong? Maybe not everyone wants to move as quickly as the early adopters?
It’s worth considering, but… I think we can interpret this differently. It might be true that people are overwhelmed by all the technology in their lives.
However, everyone adopts something new that they use as a benchmark for what IT needs to deliver. I might use multiple devices and enjoy screen size optimization while my wife my enjoys the convenience of TouchID. In aggregate, IT needs to adopt both, even though each of us uses new features at a slower pace.
A problem of scale exaggerated by richer personas and new platforms
With the size of the market, and breadth of the way technology is infused into our lives, I think IT is in aggregate going to need to move at a scale greater than then anyone thinks possible. Even if we consider that individuals are not adopting the full spectrum of new capabilities with each release.
This makes perfect sense when we think about persona development. With all the data we have, persona’s can be more individualized. Richer. More nuanced. And that means more consideration of individualized features that may be relevant to each persona.
It also makes sense when we consider that the platform informs the experience (and digital transformation is about transforming from access to experience). Each platform (phone, tablet, watch, TV, car) is going to require specialized development to capture the specific benefits of the platform (watch: notifications, tablet: larger screen sizes, TV: family interactions, etc).
Creating a software factory with our values embedded in the platform
The answer to this is simple in my mind’s eye but complex in reality.
Companies move this slow because of control. They have to control everything to make sure they guarantees security, privacy, quality, and governance. The thing is, to ensure those values, they control the whole end-to-end development process.
In the future, IT is going to have to figure out how to have more nuanced control. IT will retain control over their critical values, but not over the end-to-end development process.
This is an incredibly demanding change I’m expecting of IT, and frankly, most probably aren’t up to the task. Which is why so many companies think their industries be disrupted by digital transformation.
Coincidentally to reading Jan’s article this morning, I saw ‘trapped in a system’ by Fred Wilson. It’s another great way to describe what we’re experiencing in technology delivered by large enterprise companies. Individual behavior is reflecting reality, but our behavior as a group reflects our own habits and interests.