A recent Chilean film called NO narrates the campaign to defeat General Pinochet’s 1988 referendum that would have allowed him to remain as President after having ruled Chile as a despot for 12 years. Despite his tight grip on power and oppressive, dictatorial rule, Pinochet was put under international pressure to allow a vote. He firmly expected to win, counting on fear and apathy among the voters for a low turnout, and strong-arm tactics and cheating at the ballot box, if necessary.

The opposition to Pinochet was faced with the dilemma of how to win despite the cards being stacked against them. They needed to convince people to vote no. Their leadership was full of downtrodden miserable politicians who had been suppressed and even tortured under Pinochet. They wanted to mount a campaign revealing the misery and injustice the dictator had caused them. But one opposition leader had his doubts and decided to turn to a marketing agency with a radically different idea.

A young ‘madman’ at the agency produced a brilliant campaign that concentrated solely on images of the bright things the future might hold by voting ‘No’. By the end of the campaign, the Chileans had roused themselves from their resignation, rallied around the colourful marketing images, slogans and song, and trounced Pinochet in the voting, to the point where the Army abandoned him as a lost cause.

As compelling as the film was, it was nowhere on my mind when I received an invitation to a breakfast briefing resembling a 1960’s agitprop poster, sporting the slogan Viva la Revolucion. I dismissed the imagery as irrelevant hyperbole. The invitation came from Coverity and SQS. The briefing was about Development Testing and as I am interested in Development Testing and breakfast, I accepted, even though breakfast was to be interrupted by, ‘revolutionary speech and crowd response’. Would there be a bun fight, I wondered?

The revolutionary theme stemmed from the notion that testing is undergoing revolutionary change, with a shift in emphasis from late-cycle to development stage testing. The briefing had three participants:

  • Coverity who uses Development Testing as a ‘theme’ around which to market a set of development stage static analysis tools,
  • SQS, a service provider traditionally in the late-cycle testing space, now partnering with Coverity to provide development testing services, and
  • Aviva, a joint customer in the throes of the revolution.

The presentations – essentially by the Aviva managers – were suitably low-key: there was no hard sell. In fact, it was hard to know except in the loosest way how Coverity and even more so, SQS, figured in the Aviva story.

Aviva’s story is that of a typical large financial institution needing to modernise its development processes to be more responsive, agile and innovative. In other words, to deliver faster and more efficiently. Their journey is to overhaul their end-to-end development process, and development testing is only part of that. Streamlining development processes like source management, builds, integration and delivery – and the automation of much of that – was really the narrative. Use of Coverity’s products in development figured as they shift some error detection away from late-cycle ‘system’ testing to the development stage.

In the Q&A, it became clear that Aviva’s revolutionary road still had some way to go. The organisation is still grappling with how to shift testers (never mind the testing) out of their late-cycle comfort zone back into development. Agile practices are part of that move, but developers are still not predominantly responsible for their unit tests. And the use of Coverity analysis tools is still the subject of a period of developers learning to accept them (much of the usual talk about minimising ‘false positives’ from static analysis.) So the revolution might be started but it is certainly not yet won.

At the end of briefing (after the agitated masses had mostly dispersed), I managed to speak directly to Ben Riches, head of application design and delivery at Aviva UK’s General Insurance (GI) division, about the ‘selling process’ needed to gain acceptance and take the transformation to its conclusion. He explained how he was doing what I consider to be the usual practical things for change management: establish clear goals, get management buy-in, do education and training, give high visibility to achievements, encourage light competition among teams, find local champions, do progressive rollouts, run pilots, etc. All sound stuff.

There was still one specific point I was curious about. Given that the Coverity story is rooted in static analysis, why had I heard nothing about one of the linchpins for introducing static analysis: reduction of technical debt? As I had outlined in my paper, Static analysis for business agility, technical debt is the source of a lot of the friction that stops IT responsiveness. Was it not a problem for Aviva?

The answer could have been straight from the mouth of the adman in the film, NO. According to Ben, technical debt is just a reminder of their past misery and mistakes. The only way to sell the change – to the whole development process, not just development testing – is through painting a positive vision of what will be possible in the future: a responsive, effective IT organisation capable of defending and enhancing Aviva’s competitiveness. Will Aviva have a clear-cut win? We don’t know. The revolution has still to take its full course.