Sparking the next cycle of IT spending

Who, in the entire IT space, wouldn’t like to see an uptick in tech spending?  Enterprises would see new purchases easier to make, vendors would make more money, and technologists in general would have a new sense of excitement and mission.  It seems like we’ve been stuck in a do-more-for-less rut, but the past offers us some evidence of how we could get out of it.

If you were to plot of the growth in enterprise IT spending versus GDP growth for the US over the entire life of information technology, you’d see not a hockey stick but a series of peaks and valleys.  You would see that there are three clear periods or cycles where IT spending has significantly outstripped GDP growth, and that we’ve been in a trough ever since the last one ended in about 2000.  We’ve never had two decades pass without another cycle, so what’s wrong?  Answer: Nothing’s driving one now.

If we look at the last three cycles, they align with phases in the evolution of how IT supports worker empowerment.  In the “mainframe” phase, we applied computer technology to the creation of business reports and management information.  In the “online transaction processing” phase, we extended IT closer to the worker, and the work, and the “personal computing” phase continued that extension.  Looking at it this way, it seems clear that it’s the ability of IT to present new productivity and empowerment options that launches a cycle.  Get IT closer to the worker, and it builds a new business case for new spending.

You might wonder why smartphones and mobility haven’t launched another IT cycle, and the answer is that we tend to use mobile devices as portable desktops.  Having your information appliance in your hand may make it closer to you, but it doesn’t necessarily make it closer to your work.  What seems likely to be the Next Great Step in cycles and empowerment is point-of-activity empowerment.  Instead of providing workers with information on a laptop/desktop or on a mobile screen, we integrate IT through mobile devices directly with our work practices themselves.

The barrier to this is knowing what the work practices are, what the worker needs and when they need it.  We need to present information to workers in context, which is why I think of the next cycle driver as being contextual computing.  It’s a combination of changes to IT, widespread use of IoT information, artificial intelligence and machine learning, and cloud computing.  That fact is both the source of the strength of the idea, and its greatest weakness.

To make contextual computing work, we need to decompose “jobs” into “tasks”, and we need to be able to map the information needs of workers to these specific tasks.  This might mean some work from Enterprise Architects in framing business processes for interactive IT assistance, versus having users go to their IT devices and framing their tasks around the results.

We also need to be able to track workers, not only within the task structures, but by their location.  It’s easy to see how location tracking could work; GPS is available on virtually every mobile device, and IoT sensors could give us precise knowledge of worker movements within a facility like a tank farm or warehouse.  Tracking task progress could be done by monitoring the output of the worker, perhaps by mapping the expected versus actual application accesses performed.

In effect, what contextual computing is doing is creating a digital twin, an alternate reality that software can monitor and control, and that is linked to the workers’ real world so that what’s done and seen in one is repeated in the other.  It’s almost certain that creating and sustaining the digital twin is an artificial intelligence and machine learning application.  It’s this digital mirroring capability that lets us inject IT support into everything a worker has to do, via whatever information portal or portals are available, and AI/ML is the best way to coordinate it.

If all this sounds too good to be true, it’s because it might be.  Think of all the pieces that are needed, and all the players whose cooperative efforts would have to provide them.  There’s nothing new about this concept; I’ve written about it for half a decade at least, and I did a global financial analyst presentation of the cycles concept almost 20 years ago.  We still don’t have that cooperative synthesis, and the big question for contextual computing, and for driving another positive IT cycle, is how we could get it.

The past cycles were driven by vendors, by a shift in technology that created a new productivity paradigm.  In fact, we could fairly say that every past cycle was driven by a single vendor, IBM.  They popularized and drove commercial computing in the first cyclic wave.  Their transaction processing capability (CICS, for those interested in the nitty gritty) launched the second wave, and their PC the third.  Can IBM, or any vendor, pull together the pieces we need?

These days, for something like contextual computing, we’re probably reliant on open-source software initiatives, and I’m a believer in them, with a qualification.  The best open-source contextual computing solution wouldn’t likely be developed by a community.  It would be done by a single player and supported and enhanced by a community.  Open-source simplifies the challenge that being the seminal vendor for contextual computing would pose, but some vendor is still going to have to build the framework and make that initial submission.  Otherwise we’ll spend years and won’t accomplish anything in the end.

IT changes have driven every cycle so far, and network vendors are unlikely to play a key role in creating another cycle, because all the work really has to be done at the business application and software-platform level.  Among the IT vendors, we could see something coming out of Dell, HPE, IBM/Red Hat, or VMware, because all of them have solid IT platforms, open-source experience, and people who could likely grasp the problem.  Most important of all, each of these vendors could hope to get a nice piece of that next-cycle spending.

Past cycles have raised IT spending growth, on the average, 40% higher than GDP growth.  They’ve been responsible for all the golden ages that those of us (like me) who have seen many or most of those cycles remember.  I’d like to see another cycle, and so I’d like to see a vendor, some vendor, step up and start the next cycle rolling.  Takers, anyone?