Back to blog

The Language of Talking to Developers: The Importance of Outcome-Based Segmentation

Christina Voskoglou · November 27, 2013

Why outcome-based segmentation should be the cornerstone of developer outreach strategies. VisionMobile’s Data and Operations Manager, Christina Voskoglou, explains why everyone running a developer program should focus on outcome-based segmentation and not technologies, demographics and platforms.

Blog006_Final_Web

00 Shooting the duck

Two statisticians were hunting for ducks by a creek. Spotting one taking off behind a bush both hunters fired simultaneously: The first man’s shot fell 1 meter too low while his friend’s shot flew 1 meter too high. Thrilled, they dropped their rifles and started congratulating each other, hopping about, happily chanting: “We got it on average, we got it on average!”. The duck, even happier than the statisticians, had of course in the meanwhile flown away to safety.

This is a lesson on how working with averages is a sure-fire strategy to miss your targets. [tweetable]There is no average person. And there’s no average developer[/tweetable].

01 “Are you talkin’ to me?”

There are mobile platform companies like Microsoft and Apple. There are ad networks like AdMob and Inneractive. There are back-end tools companies like Parse and StackMob. There are cross platform tools like Appcelerator and PhoneGap. All of them are serving developers. All of them are competing for developer attention with marketing dollars. Most of them are talking to the Average Developer.

Devs_currency_of_ecosys

Consumer segmentation has been a popular strategy since airlines figured out how to sell an airplane seat in 100 different ways to 100 different people. Segmenting a population to better understand it has long since been proven to be doing a better job at getting those campaign response rates up than simply addressing the whole population with an ‘on-average’ interesting message. Now companies from Amazon to ZTE are figuring out how to grab developer attention through segmentation.

The bigger and more heterogeneous a population is, the more effective segment-based outreach campaigns are. And one could hardly wish for a faster-growing and more diverse population than mobile developers across geographies and platforms. Question is, how do you segment developers effectively to make sure you’re talking to the right people with the right message?

The classic approach practiced by platform vendors is to segment based on demographics, skills or technology considerations such as platforms – e.g. Android vs. iOS developers. Trouble is, technology used is a choice that may have been made by different developers for different reasons. For example, a developer may have chosen Android to take advantage of its wide customer base, while others because of its lower cost of development vs. iOS. Simply grouping together people who made the same choice doesn’t necessarily result in a homogeneous group that behaves in a coherent way – choices made by Android developers who care more about reach will differ to choices made by Android developers who want to learn at low cost. If we fail to understand the drivers behind these choices we won’t be able to understand how and why choices change over time, for example why developers would switch from native to HTML5 or switch their main platform from Android to Windows Phone. Nor will we be able to see how to convince developers to switch from an existing choice to a new one – e.g. from a tool they use today to a new one offered – or why and how developers decide to commit resources and take risks in the app economy.

[tweetable]Effective developer segmentation can only be based on what drives developers[/tweetable], what they’re trying to achieve in the app economy. In our Developer Segmentation report we adopted an outcome-based approach to segmentation, backed by hard data from a survey of 6,000+ app developers. The research resulted in eight distinct behavioural segments: The Hobbyists, the Explorers, the Hunters, the Guns for Hire, the Product Extenders, the Digital Content Managers, the Gold Seekers and the Enterprise IT. By focusing on the motivations that drive developers to adopt a new technology our model provides actionable insights on which developer groups to approach and how.

03 A success story

Let’s consider a common scenario of an outreach program pivoted around supporting developers to succeed in the app economy. So, how do developers succeed in the app economy – how do they perceive success? You can simply assume that most people care about revenues – for example Microsoft and Nokia’s AppCampus program is pivoted on rewarding developers with money for a limited app exclusivity.

However, our research proves that [tweetable]revenue-based success is only half the truth[/tweetable]: We found that just above 50% of all developers give success a definition that involves direct revenues. This implies that if you addressed all developers out there with a revenues-oriented message, about half your marketing budget would be predictably wasted.

What if you segmented developers based on the platforms they use – Android vs. iOS vs. Windows Phone et-al? The next chart presents the popularity of four success metrics (out of the six explored in our research in total) within each platform – Android, BlackBerry 10, HTML5 mobile, Windows Phone and iOS. We have normalised the results over the developers’ desired outcomes in the app economy to remove bias and isolate the effect that platform choice alone has on success definitions.

Platform-defined segments

Differences between platform-defined segments are obviously very small to be actionable. For example, there is no great difference in what platform segments think about reach as a success metric – around 60% of all segments alike say reach is important. Platform-defined segments obviously fail to define distinct clusters of developers that perceive success in different ways. If you’ve used platform-based segmentation to understand your developers, you’ve been wasting your money.

Segmenting developers based on app categories leads to similarly un-actionable results. For example take four category-based developer segments – those developing games, apps, games and apps and enterprise apps. How do developers differ based on their success metrics? Do game developers vs. app developers measure success in different ways?
Again, technology-based segmentation fails to understand differences among developers with respect to success perceptions: All four app-category segments give approximately the same weight to knowledge gained (47%), while exactly the same percentage within each segment (6%) doesn’t care about success at all. ‘Games only’ developers seem to differentiate themselves from the other three segments with respect to reach and slightly with respect to direct revenues, however the resulting differences are not intuitive: Why would game developers care less about reach and more about revenues than other segments?

Success Perceptions

04 What are you trying to achieve?

In our Developer Segmentation research, we segment developers in terms of what developers are trying to achieve. This is based on the Jobs-To-Be-Done segmentation methodology popularised by Harvard professor Clay Christensen. The results are not just refreshingly clear. They are refreshingly actionable.

Consider the success metrics distribution for five of our eight outcome-based segments as shown in the graph below. First, notice how direct revenues are important to only 27% of Hobbyists as compared to 81% of Hunters. Hobbyists more than anyone else state that they don’t care about success (23%), while Explorers define it mostly in terms of knowledge gained (72%). Product Extenders, who aim in promoting a non-mobile product, care most about reach (70%).

Measure Success

It is evident that [tweetable]outcome-based segmentation results in distinct behavioural groups[/tweetable]. Success metrics is obviously not the only instance where segment dissimilarity occurs. In our Developer Segmentation report we explore how outcome-based segments differ in their motivations (personal, commercial, community), in the choices they make (including platforms and tools), in the way they think (challenges, platform selection reasons), in markets they target (devices, audiences, app categories) and in the ways they make money (revenue models, monthly revenues made). We also profile them in detail (including their geographical distribution, their experience and role) and discuss what share of the app economy each segment holds.

05 Mind the sub

It may be that you are interested in understanding a subpopulation of developers – you may be addressing only developers of a particular region, or developers that use a particular platform. In such cases, you will get more accurate profiles if you segment only the subpopulation of particular interest – e.g. Asian or Android developers – rather than the whole developer population.

Segments of regional subpopulations

Consider for example Hunters in Russia vs. those in the US. As Hunters will always be hunters, both groups are mainly after revenues. However, Russian Hunters care more than their North American colleagues about the knowledge they gain, and also a slightly higher percentage states to be indifferent to success. These differences point to a less ‘mature’ market and therefore a less ‘mature’ set of hunters in Russia: Most probably a higher percentage of Hunters there than in the US are at their early stages, currently more focused on building know-how rather than reaching customers – something you wouldn’t have picked about Russian developers if you had just used the ‘global’ profile of Hunters.

If you care to understand developers, you need to understand what motivates them, what they consider success and what they’re trying to achieve. Our Developer Segmentation research has done just that.

What other stories have you seen where developers are poorly understood or mis-marketed?

– Christina (@ChristinaVoskog)