Managing device heterogeneity in enterprise mobile services

9 May 2006 by Rod McLaren

Recently we compared the new BlackBerry 8700 to its parents, the 7730 and 7100/30. In many respects, these changes are unambiguous improvements, certainly for individual users who request the upgrade from their mobile operator. (Though it has to be said that early sales figures aren’t great.)

But what happens when the upgrade isn’t requested? This situation can arise for companies who run a fleet of mobile devices: when an existing device dies the mobile operator may replace it with a newer model if it’s a cheaper option than replacing it with another of the older model or repairing it.

Despite the new device being better than the old and the replacement being offered at a good deal, the result can often be uncertainty and nervousness for the customer: will our operation (or people and processes) be able to work happily with the new device? Will we need to retrain users? Will the existing applications work on the new device?

This recently happened on a service we supplied to a customer. The solution is a device and an application running on it for users in the field, two different web front ends for users at regional locations and head-office/data centres, and a large amount of back-end glue and data infrastructure. It has been running for 3+ years, and the customer has spent huge efforts in successfully bedding it into their organisation. The users are trained and running on the existing system – this is particularly relevant for the users in the field who work in testing conditions, and is the reason that device heterogeneity isn’t simply managed by developing for a generalised platform and device form-factor.

Presented with this situation, the customer has a tricky choice to make:

1. Find more of the obsoleted component. This is likely to be a temporary solution at best, but can be a useful stopgap.
2. Accept the introduction of device heterogeneity into the operation, and embrace it. This places an immediate burden on testing the technology and applications to ensure they work consistently across the device range, and on retraining users.
3. Replace everything with the new device to rigorously preserve systemic/operational homogeneity. Again, a similar burden on technology and training, though in comparison the the cost is likely to be larger in the immediate term and smaller long-term.

Five actions vendors and customers can take to manage this situation better:

1. Stockpile existing devices. This allows the customer more time to prepare technology and retraining, but they have to accept the increased capital investment it requires.
2. Design into the service lifecycle a transition plan of technology and training, and regularly review the plan. And accept the management overhead that comes with it. For example, bind it into maintenance agreements.
3. Design every component of the system with potential obsolescence or replaceability in mind. For vendors, this may require them to be more honest about how the customer could replace them, though this honesty can often have the positive effect of increasing the customer’s loyalty to the vendor.
4. Design applications to be more multi-platform-ready – because a vendor can’t tell what (or what kind of) device the mobile operator will push to the customer – or the customer will want to switch to. But there’s a balance to strike: the development cost may increase, and the applications may be less user-friendly if they don’t take advantages of device-specific benefits. And because the developer cannot predict which devices will be in the market in years to come, there will be an irreduceable testing burden as new devices are rolled out to an existing system.
5. Be careful of relying on solution channels where the vendor cares about only one piece of your solution. In this case, the mobile operator understandably cares more about the customer’s subscription to their pipe than it does about device consistency.

Some of these actions are also useful in situations where change isn’t driven by new devices, eg where there are contractually- or compliance-driven changes to systems and services.

Related:
Daniel Taylor’s posts at mobile Enterprise are relevant here. In Addressing the ‘Channel Gap’ for Enterprise Mobility and CTIA: Which Channel Are We Talking About? he makes the case for a traditional three-tiered sales channel to enterprise – with Value-Added Resellers and System Integrators in the mix (particularly for what he calls the “last mile” to the enterprise, and to go beyond the mobile email model) as well as the mobile operators we see today.

Werner Vogels (Amazon CTO) on heterogeneity as caused by system scale-out :
“A second problem area [with scalability] is that growing a system through scale-out generally results in a system that has to come to terms with heterogeneity. Resources in the system increase in diversity as next generations of hardware come on line, as bigger or more powerful resources become more cost-effective or when some resources are placed further apart. Heterogeneity means that some nodes will be able to process faster or store more data than other nodes in a system and algorithms that rely on uniformity either break down under these conditions or underutilize the newer resources. [...] For the systems we build we must carefully inspect along which axis we expect the system to grow, where redundancy is required, and how one should handle heterogeneity in this system.”

1. Hi Rob and readers of the Mobbu blogsite

   Great posting ! (and congrats on being the lead story in this week’s Carnival of the Mobilists)

   A quick comment to help put the urgency in this.

   Corporations tend to be prepared for PC upgrades and think typically in a 3-4 year replacement cycle for those, perhaps a bit faster for laptops.

   With mobile phones the global average replacement cycle is down to 18 months, and obviously for most business users with employee-provided mobile phones, they tend to have two phones (20% of all Europeans have two phones and this is already 13% for North Americans, and over 30% in leading countries like Hong Kong, Taiwan, Israel and Italy)

   With multiple subscriptions two problems emerge for corporate mobile environments – one is that the employees may be tempted to substitute at times their other phone – note their replacement cycles are typically not in synch, are at opposite cycles in fact, so the work phone may be the new phone for a year, but then we get a new personal phone, which can easily be a superior smartphone/cameraphone than that provided by our employer

   Secondly – even more relevantly – with the replacement cycles shrinking, all handset makers are shrinking the full life cycles of their model ranges….

   So your advice such as stockpiling becomes even more critical.

   EXCELLENT POSTING !!

   I also maintain a blogsite on IT/tech and marketing, often blogging about mobile, so your readers might want to visit at www.communities-dominate.blogs.com

   Obviously my four bestselling books on 3G mobile and digital convergence touch on these topics in more detail. More of those at my blogsite and my website.

   Keep up the good work here at Mobbu !

   Tomi T Ahonen
   4-time bestselling author and 3G strategy consultant
   www.tomiahonen.com

   
   — Tomi T Ahonen    19 May 2006, 16:56    #

2. Hi Tomi, I was only thinking about change to corporate/enterprise programmes as driven by operators, so your point that it’s equally driven by consumer (employee) demand – at least for simple apps like mobile email – is indeed urgent and relevant.

   Stockpiling devices can only ever a temporary measure, unless a corporate is prepared to carry a lot of inventory at the back of the IT dept! Though it occurs to me that someone must surely have found a good business model in doing that.

   “Corporations tend to be prepared for PC upgrades and think typically in a 3-4 year replacement cycle for those, perhaps a bit faster for laptops. With mobile phones the global average replacement cycle is down to 18 months…” – I would imagine that the other problem is that some enterprises have little or no replacement cycle plan at all (yet), particularly as mobile/mobile apps often enter businesses in a very tactical way.

   Say hi to Alan Moore – I know him from tv days with John Nolan et al.

   
   — Rod McLaren    19 May 2006, 19:02    #

BlackBerry 8700 and 7730 comparison The Carnival of the Mobilists gets its own home