Thought Leadership

Recently a couple of our clients have wanted to test their standing in the minds of developers as to their perceived “leadership” in various areas. It’s an important subject beacause it actually measures your credibility with your target audience, and credibility is critical to how well your marketing efforts work. The perceived leader is already ahead of the competition before even one positioning statement is made or one promotional program is begun, Consequently, if you know where you stand in the minds of your audience, even if you’re not number one, you have valuable knowledge that can be used to tailor your messages and advance your leadership standing. Some of these clients refer to this as “thought leadership”.

Of course, thought leadership really means leadership in the thoughts of your audience and that leadership can apply to many things. For example, your company may be perceived of as the leader in providing cutting edge technology but also be perceived of as a laggard in customer service. Company A might be the most “trustworthy” but Company B has the “Most potential for the future” The fact that there are many facets wherein your company might be the leader also helps with positioning and messaging. Pick the course of least resistance as a lead in to your audience and then build on that to enhance the areas where you are seen as weak. It’s a powerful marketing strategy.

We’ve devised a Thought Leadership measurement service using 10 different attributes of relative market positioning and an additional 12 psychographic metrics to help clients understand where they fall in the landscape of the industry and their competition. We can pinpoint your position in the minds of developers and suggest ways using the resulting information to approach your audience most effectively and to emphasize perceived strengths as well as address perceived weaknesses. Perceptions are everything when it comes to marketing. See if we can help you.

Projecting Developer Population

Twice a year we publish a study called Global Developer Population and Demographics Study. It projects the developer population worldwide by region and by major countries as well as layering global survey data from our twice yearly Global Development Survey to provide estimates of how many developers there are and what they’re doing. Projecting populations is easy for countries like the U.S. who publish census statistics by job codes – the only trick is deciding which codes fit into your definition of a developer. But very few countries publish these statistics, so how do we estimate the populations in most of the rest of the countries in the world?

We spend considerable time researching what statistics are available from organizations such as World Bank, UNDP, CIA, and the governments of the countries themselves. We use country population, GDP, tertiary education, and whatever other technical markers are available for that country such as computers sold, mobile phones shipped, internet connections etc. From there we can make an estimate by applying a formula derived from what we can see in the US as to how these statistics roll up to deliver a developer. Growth rates are derived from linear regressions modified for real world events. Since we’ve been publishing this report twice a year since 2006, and are continually monitoring it, we can show some solid history with reliable forecasts.

Since the results always depend on what variables are put into the mix, it’s very important that these are relevant as well as reliable. One variable that is no longer relevant and one that can be misleading is developer spending. We know of one firm that uses this as a variable in calculating the developer population of countries with no published job census data. While this used to be relevant, the easy availability of open source and free tools has made developer spending into a declining variable, even though the actual population of developers has increased. This leads to crazy results, such as estimating a population for 2010 that is half of what the firm estimated in 2007.

Knowing developer populations is important for sales as well as product and program strategy. If you have questions about developer population and would like to talk to an analyst – give us a call.

The Ultrabooks have Arrived

….and just in time, I’d say. For anyone who wanted the slim, easy to deal with style of the Mac Air but didn’t want to go down the Apple / Mac OS route, it’s been frustrating. But now a slew of sexy svelte designs are coming to market with incredible performance and battery endurance powered by Intel Core i3, 5 or 7 and optimized by their partners. Today both Toshiba and Lenova launched their ultrabooks and more are on the way this month from Samsung, Acer and Asus.

These new laptops were shown recently at Intel’s Software Analyst day along with presentations higlighting some very impressive innovations and advances in performance and power optimization. For example, how about battery power that will last 12 hours while fully engaged and up to 3 weeks while on standby? Couple that with the efficiency and performance of the Intel multi-core chipset and onboard HD 3000 graphics, all built into some very sleek form factors and suddenly there’s real competition in town.

I’m usually a dinosaur when it comes to upgrading my computers since it’s such just a hassle, but these new laptops add some real motivation – I think even I am going to upgrade this Fall.

Dynamic NPS for Competitive Advantage

The Net Promoter Score (NPS) was created by Satmetrix, Bain & Company, and Fred Reichheld to answer the “ultimate question” of whether or not your customers would recommend you. It asks the respondent to gauge on a scale of one to ten the likelihood of them recommending your company (or service or product). You can then calculate a score and benchmark against it.

If repeated at regular intervals through a survey of your user base this gives a good metric on which to measure the succes of your programs, marketing campigns or other initiatives among your existing customers. However it gives no inisght into how your score stacks up to other products or programs. There is no way to know from the score alone where you stand amongst the competition. And the frustrating thing is that a general survey of the population (and not just your users) gives skewed results since those respondents who may be unfamiliar with your offerings will more than likely give you either a zero (meaning no info) or a five (which is directly in the middle and thus considered neutral by most people) but both of those scores fall firmly inthe detractor category and alter your overall score.

We’ve got that problem solved. What we’ve been doing for our clients is to identify four or five major competitors and then run a dynamic survey in which only developers who are familiar with each of the competing companys or company offerings are asked to rate that company. In that way we can provide an NPS score for your company as well as NPS scores for your major competitors and all of the scores come from developers who have enough info to be valid participants. We can also look at the socres according to a different categorization of the scale so that you can see the original score plus a new variation that might make more sense. You can have both.

In addition we dynamically ask why each respondent answered the way he did to provide actionable data. You’ll know your NPS score, your competitors NPS scores and what to do to raise your own score. Repeat in six months or a year (depending on how fast your organization can move) to see how you’ve improved.

Mobility in the Enterprise

The enterprise client is morphing faster than many IT managers can keep up with. Mobile devices are already a common target for corporate enterprises according to our latest Mobile Development survey. Seventy-four percent of respondent developers said they would be extending enterprise apps to mobile clients in the coming year. The survey, which was conducted in June and examines the technology adoption and intentions of over 400 mobile developers, also found that Cloud or SaaS enterprise APIs were one of the four top API types in demand by developers.

Why is this important? It means that the corporate enterprise is fragmenting in its client targets. Employees are often using their own smartphones and expecting to be able to access in-house apps or data from the road, from their homes etc. Soon they’ll all be carrying tablets, and it’s not likely that any one tablet will dominate the market, so that will just add to the issues that corporate developers need to face when figuring out which clients to support.

This may be the driver that precipitates the widespread adoption of web apps over native apps for mobile development, at least in the corporate world. While performance and some functionality is sacrificed when running an app through a browser, cross-platform capabilities shine, and that might turn out to be the most important thing of all.

Changing Landscape

Changing Landscape

We’ve been monitoring technology adoption patterns here at Evans Data Corp for more than a dozen years, and I have to say it’s always interesting. Sometimes the changes are subtle and slip easily into place without much fanfare, but sometimes the changes are monumental. This is a time when the landscape is changing dramatically and it’s fascinating to watch as developers embrace different technologies to adapt.

Language use patterns have turned on their heads in the last year with Java use up in all regions of the world. This is due to the intense focus on the mobile client, and is particularly true in Asia where developers have always had more mobile focus. But there’s is also a resurgence of the use of C++ and we only can speculate at this time (we’ll know more after the Fall survey) but it may correlate with a perceived need for better performance in a variety of apps including apps for multiple processors or cores.

What about data technology? After so many years of ubiquitous SQL use we’re now seeing a lot of interest in NoSQL (Not Only SQL) and the reason is the new availablity of “Big Data” – and new types of data that don’t fit nicely into a relational database. And the interesting thing about the curiously quick adoption is that it’s being embraced by enterprise developers much more strongly than any other segment.

Cloud, mobile, new data types, and other seismic computing changes are shaking up the development landscape and its now hitting the corporate enterprise. Any vendor serious about playing in the enterprise needs to become aware of these changes and find an answer for them.

C++ and Java Resurgence

The latest edition of our Global Development survey series is just about to ship – comprised of three different reports: one on North America, one on EMEA and the other focused on APAC development. While the different geographies usually return results that are different, in this case we’ve seen a thread that runs across all three – the seeming resurgence of both Java and C/C++. This is surprising in that the number of users of either language has been declining in recent years to show a steady trend as the use of scripting languages presents the opposite curve. But this Spring we see increased use of C/C++ and Java across regions.

Makes you wonder why this might be so. Think about C++ for instance. This is a serious language that lets developers do things they wouldn’t be able to do in a more managed language. Adding restrictions in a more modern language like C# provides safeguards but also limits flexibility. Are developers just getting tired of being looked after and yearn for the old cowboy days when C++ dominated and the sky was the limit to what they could do? Or maybe it’s something more practical. C++ allows for better performance in most cases – important in games, imaging, and many scientific applications, and these areas are growing. Could that be it?

With Java the resurgence is almost certainly due to the huge growth in smartphones as a target. Java is perfectly suited to Android, Blackberry, Symbian etc. It provides an element of cross-platform capability that’s extremely attractive to developers in the fractured mobile landscape. With mobile devices replacing PCs as the dominant target, Java’s got a great future.

Programming languages come and go based on the needs of developers trying to stay on top of the demands of the market. I think what we’re seeing here is a marketplace that’s evolving into one that favors both of these languages. C/C++ and Java are coming back!

Intel’s Impressive Development Tools

Intel’s innovations are sometimes described as the tick tock model where each tick is a new manufacturing technology and each tock is a new micro-architecture. But that just describes their chipsets. It doesn’t begin to describe the continual evolution of their software development products which have now blossomed into some of the most impressive tools in the industry.

I’m just back from the Intel ISTEP conference held this year in Dubrovnik, where Intel not only showed off a clear and coherent vision for a robust family of software development tools but also put them through their paces exhibiting some of the most intuitive UIs for performance tuning and multi-threaded development that I have seen coupled with powerful capabilities and great flexibility. Take Parallel Building Blocks, for example. Intel has been publishing and promoting tools for multi-threaded development for years and was largely responsible for bringing parallelism to the “masses” via Parallel Studio. But developers often need different tools and techniques for achieving parallelism in differing circumstances and that’s where parallel Building Blocks comes in. It consists of three components for three programming models that can be used separately or in combination. There’s Cilk Plus – the jewel they perfected from their acquisition of Cilk Arts about a year ago. It provides language extensions for C and C++ to enable task and vector parallelism, and works with Windows or Linux to easily enable parallelism. The second building block is the venerable Threading Building Blocks template library for C++ which has been aiding developers in threading their apps for years, but keeps on getting better – it now supports most major operating systems, including Mac OS. Finally Array Building Blocks addresses data parallelism through a sophisticated C++ template. Together the three provide for almost any circumstance.

It used to be that Intel and software products wasn’t a match that immediately came to mind, but these development tools are changing all that. As parallelism becomes more and more imperative for the creation of efficient apps, the Intel family of tools is becoming both more important to developers and way more impressive.

Agile Ops

The perennial tension between the development team and operations may be getting closer to easing with the introduction by IBM of the concept of Agile Opps.  Similar to the concept of Devopps but more robust and planned out, this is an approach that seeks to promote a closer interaction between the two areas and hence a more streamlined process with less aggravation on both sides and less problems stemming from a “siloed” approach.

But let’s step back a minute and examine the problem.  From the operational side, the development team often seems to live in a world of their own, just occassionally tossing programs over the wall that the operations team has to set up and deploy  – often in heterogenous situations and often with difficulty.

The development team, on the other hand, received requirements for the software and often several change mandates from management, is responsible for architecting, coding, testing, and debugging the software, feels  pride in its creation and often has very little sympathy when operations “breaks” their baby.  In addition, the majority of developers today embrace Agile programming in which there is frequent interaction with stakeholders and many of them feel that should be sufficient to assure that the concerns of operations are adequately addressed.

What IBM is doing with Agile Opps is to pull together a set of best practices and possibly new tools from both their Rational and Tivoli teams, representing respectively their development tools group and their operations product teams.  The result sounds very promising as these two groups are now working much more closely and each brings its own particular point of view and insights.  Tivoli has recently benefited from an injection of talent transferred from Rational which will only deepen the synergy.  While Agile Opps is still a work in progress, it has an excellent start and great potential to create more harmony between the development and operations of their customers and so deliver a competitive advantage.

Not So Timely

Our most recent Global Development survey series showed that fifty-one percent of North American developers say their latest project was delivered on time, and two-thirds of their projects came in within 3 weeks of target date. That might not seem so bad to people who have been active as product managers or development managers. Fact is it’s hard to predict how much time a software project is going to take. How can you know how the bugs will be injected, where they’ll be and how long it will take to find them? Software development is in many ways more an art than a science.

Asian developers were slightly less likely to be on time, but 71% came in within 3 weeks, while European developers delivered only 36% of their projects on time, and less than 60% were delivered within 3 weeks of target dates. Those most likely to be on time were developers writing internal corporate apps for the enterprise.

Sometimes the projects were not just late but discontinued altogether – the ultimate form of late. The phase during which projects were most likely to be discontinued was the coding phase in both North America and EMEA, but in the APAC region things work a little differently. Projects there were more likely to be abandoned during external beta testing. If we consider these situations, the Asian one sounds much worse. Aftter all, how discouraging to abandon the project after it’s already in external testing. Does this imply that the bugs they found in beta were too bad to try to fix – or is it that the app itself was so uninspiring that users couldn’t even get through testing?

Whichever, some of the code generated even in these projects can be used infuture porjects, and much of it is. That’s the magic of reusable code.

Now it’s on to the magic of Christmas. Have a great holiday!