Surfing the mobile shift: Where are all the developers?

Surfing the mobile shift: Where are all the developers?

Summary: Are we a mobile first society? And, if so, how are developers keeping pace with the shift? Looking at recent data, Jeffrey Hammond is worried that development teams are not prepared .


Ever hear about the myth of the “seventh wave”?  Surfers use it to describe the big one -- the wave that you can ride all the way into the beach.

While it’s been a while since I’ve tested its premise at the shore, I often think about the seventh wave when dealing with the constant waves of tools, processes and technology we developers face. With the constant change we face, how do you determine which technologies will change everything from over-hyped vendor pabulum (3-D TV anyone?)

We don’t have the capability to invest in every new technical advance that comes down the pike, so we need to be able to tell the seventh wave technologies from the others that might provide incremental productivity benefits or cost reduction, but that don’t change everything we do or think.

I’ve personally seen three seventh waves in my 20+ year development career. The first was at the start of my time as a professional developer – the Client/Server wave. The second was Web 1.0 – I remember watching long established ISVs struggle to adapt to the revolution that Mosaic touched off.  Now we’re a few years into another seventh wave – the shift to “Mobile First” development.

It’s easy to miss the structural changes when you’re in the middle of it. It’s like being a gardener  – you don’t see how quickly your crops grow when you see them every day; but if you go on an extended business trip, it seem like everything grows like crazy while you’re gone.

This “step back and see the change” point was driven home to me by this data nugget from Forrester's Forrsights Workforce Employee Survey, Q4 2012, which is featured in Ted Schadler’s latest report on 2013 Mobile Workforce Adoption Trends. In particular, this graphic jumped out at me:

I was struck that the ONLY place where global information workers use a computer more than a tablet or a smartphone is when they are sitting in their cube or office! Then it hit me – we’ve made the transition to mobile first. It’s in front of our eyes, but we’re not seeing it. 

Think about how this graphic plays out in your day-to-day activities. When you walk into a conference room, how many people are using iPads vs. Android devices vs. Windows laptops? Now that I consciously think about it, the change in the past two years is substantial, but it happened so gradually that it was almost invisible.

We’re in danger of becoming this generation’s mainframe developers – still a lot to do and with a very important set of tasks, but no longer at the edge of business innovation.

Now I’m a bit worried. Our latest data indicates only ~20% of developers are actively engaged in mobile projects, while our customers have moved to a mobile first world. We’re collectively behind the crest of the wave, and I fear many of us might lose the edge and get stuck in the trough behind it.

Our development organizations will not be able to keep up with what our customers and co-workers demand. As a result, we’ll be maligned as laggards and perhaps even replaced by a new generation of modern application developers who understand what it takes to build these new systems.

We’re in danger of becoming this generation’s mainframe developers – still a lot to do and with a very important set of tasks, but no longer at the edge of business innovation.

Our development shops have to start moving faster so we can catch up with the habits of our customers and co-workers. It’s in this spirit that we’ve launched our Mobile Application Development playbook. It’s designed to help identify and overcome:

  • Development technology challenges. The number one question I get asked about mobile development is: “HTML 5 or Native”? My answer: “It depends”. When you understand the benefits and drawbacks of native, hybrid, middleware and web based development and how they match to your customer’s engagement expectations, making a technology choice becomes a lot easier.
  • Difference in development culture. Developing mobile apps is very different from the traditional systems of record your teams have been building for the past 20+ years. The technology choices are easy to make in comparison. If you’re not using Agile and dev-ops practices, continuous delivery or don’t know how to launch a minimum viable product, you’re going to struggle with mobile development. 
  • Integration challenges. I’ve seen a lot of client that have implemented a first round of mobile apps by working with a third party design agency or regional Sis. Now these apps are evolving into connected products, and they need to tie into existing system of record and system of operation. These phase two mobile apps are a lot more complicated and the business can’t simply go around IT to get them built.
  • Evolution in systems architecture. Many development teams try to port the tightly coupled, stateful, MVC-style apps they’ve written on big application servers into the world of omni-channel mobile clients. It doesn’t work very well. Modern applications are built different, scale different, and are deployed differently than what many of us are used to.
  • Evolving success metrics. While a “5-star app” is the ultimate measure of consumer success, there are also financial and productivity metrics that guide the evaluation of B2C and B2B mobile efforts. Understanding what to measure (and why) is an important part of the mobile shift.

Jeffrey Hammond is a vice president and principal analyst at Forrester Research serving Application Development & Delivery Professionals. Follow him on Twitter at @jhammond

Topics: Mobility, Software Development

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Mobile developer tools are still imature

    and there are too many incompatible platforms. Tools like Xamarin are a first step but we have a long way to go.
    • People are different, applications are different, scale different...

      These are grandeous statements. It seems to leverage the extremes of the various worlds. Yes some development shops have a very old style and process, maybe long development cycles. That wouldn't work well with rapidly changing and higher expectations of the current immature mobile world. But that's an extreme. Deployment is diffrent? Compared to what? MVC won't work? Why not? What's the alternative that is leaps and bound better? There's not much baseline with which to take this article seriously other than comparing the caricatures of 'old' and 'new'. After it matures it will likely not be very different.
      • Dismiss at your own risk

        Sure - you can use MVC to build modern apps, but that doesn't mean it's the best system pattern to use.

        I've been looking at highly rated mobile apps over the past year, and when you examine how they are built you see some real differences from traditional web apps. They push state to the edge, and use alternative system patterns like "Broker" or "Pipes and Filters". You'll notice in-memory DBMSes everywhere, Sharded SQL or noSQL DBMSes, and a bunch of other distinct characteristics. I don't think these companies set out to use completely different and new technologies - they found that they worked better, scaled better, and cost less given the particular demands of Omni-channel clients. In many IT shops, there are minimal skills WRT to these types of technologies. I'll be a bit less concerned when I start having regular conversations with IT devs that match the ones I'm having with companies that have advanced mobile strategies - and the apps and customers to go with it.


        Jeffrey S Hammond
    • Re: and there are too many incompatible platforms.

      That problem is solving itself. Linux-based OSes are gradually taking over (Android now, Ubuntu and FirefoxOS to come), and you can build for them all with a common set of tools: at the lowest level, GCC to compile native C and C++ code for ARM, MIPS or whatever architectures appear down the road; and above that, architecture-independent languages like Java+Dalvik, JavaScript+HTML5+CSS, perhaps even other possibilities like Python. All Open Source, all offering at least some level of interoperability, nobody trying to erect walls between them. That's the future of mobile.
      • I hope you're right

        I agree with many of your observations - in fact one of the projects I'm watching closely is the LLVM project ( I think is has the sort of potential to deliver what you envision. In combination w/ projects like Cordova there are cross channel ways to attack the problem.


        Jeffrey S Hammond
  • The graph appears to be misleading, pay careful attention to wording.

    "Where do you use the following devices in a typical work week?"

    Well, that's an interesting way to phrase question.

    This sounds to me like a multiple choice quiz with three questions. There's no indication of whether multiple choices were allowed, or how many people didn't fill in at least one of the choices.

    One thing is clear, though: It's *NOT* asking how many of each device people see in each place. So this is *NOT* to be confused with a survey that counts how many of each device is at each place.

    "I was struck that the ONLY place where global information workers use a computer more than a tablet or a smartphone is when they are sitting in their cube or office!"

    This is not indicating where things are being used more. Read the phrasing of the question more carefully. It's not a numerical count or a survey of devices.

    So - while I'm sure it's true that we need more mobile developers - I'd warn against using this particular graph for the purpose of making the point. It's a misleading graph.
    • What's in your cube?

      Also, what happens to be in a cube that's not anywhere else?

      Your work computer.
      • Exactly - but then there are laptops

        And what the chart up top seems to indicate that that workers are pulling out tablets more and laptops less in away from desk situations.
        Jeffrey S Hammond
    • Multiple response were allowed

      Each question did allow multiple responses - the survey was part of an annual battery we run as part of our Forrsights product.

      The responses on the last four categories did not surprise me - I would have expected them. The second and third did - what that tells me is that information workers are moving away from pulling out their laptops while in meetings, our when they are away from their desk. That's a state change from 2-3 years ago.

      What it means is that you need to make sure that applications that present information that a worker might want to consumer while in a meeting needs to look as good on a tablet as it does on a laptop (or maybe better). How many of your current applications fit the bill?


      Jeffrey S Hammond