17 August 2010
Getting Back To Outsourcing Basics
By Alexei Miller
Alexei Miller of DataArt contributes a by-lined article to Forbes Magazine. “Getting Back To Outsourcing Basics” looks at how outsourcing has become needlessly complex and how simplifying may be the recipe for success in outsourced software development.
“IT outsourcing has become the norm worldwide as a proven way to reduce costs, access global talent and drive innovation. Unlike many other mature industries, however, it has remained an area where industry-wide process standardization is almost nonexistent. Despite 25 years of trying, we are still nowhere near universal return-on-investment benchmarking, risk management or even reliable ways to measure quality.
Compare that to how buyers talk to vendors in contract electronics manufacturing, for example, and the differences are startling. In IT development, however, full standardization is impossible as services vendors often sell that which does not yet exist and, at times, that which has never been done before. Because of this, IT outsourcing remains an area where the collective experience of others does not do much to provide a basic level of trust between the buyer and vendor that their next endeavor will succeed. To mitigate the risk of failure, each new deal must work out its own relationship, its own ways to build trust and its own way of evaluating performance.
This has spawned all sorts of needlessly complex theories and best practices aimed to help, but which often only muddy the waters instead. Too often, outsourcing practitioners mistake strategy documents, dashboards and PowerPoint presentations for results. Outsourcing is often over-engineered and overly complicated. Collectively we need to get back to basics and remember that what works in outsourcing is exactly what works inside the organization--as long as one is ready to look at the two as one.
Team Dynamics and Motivation
Take staff turnover: Every buyer cringes at this metric. Frequent complaints are made about high attrition rates in emerging markets, while forgetting how competitive New York and London are. Yet turnover is simply a function of work, pay and company environment. Buyers should not assume that if they pick a piece of work their in-house staff considers boring and send it to a vendor that they will somehow consider it a privilege. If the work is boring, churn is likely to rise at the vendor just as it would with internal staff--accept that and prepare for it. But if the work is interesting and challenging--spend the time to show why--it will motivate teams when they see their participation has real impact.
Until Death Do Us Part
Strategic outsourcing relationships are sometimes compared to marriage. It's actually nothing like that, unless you are willing to accept that a successful marriage requires frequent discussion of divorce. Large-scale, multi-year deals often do more harm than good for both parties. They are difficult to negotiate, manage and get out of. They leave both sides with uncomfortable compromises and dangerous illusions; buyers concentrate risk with only one vendor, while vendors are so pressed on costs they are forced to find and allocate cheaper, subpar resources. Economies of scale are elusive in this business: A McKinsey & Co. report on offshoring models found that when outsourcing centers grew beyond 2,000 employees, the cost-performance ratio deteriorated.
Of course there are situations where only large deals are practical, but they are used far too often. Break it down, multi-source and be nimble about changing course. As an insider I know that nothing motivates vendors to deliver under budget and on time like the chance to win new work, and nothing makes them more complacent than a multi-year client commitment. There is also a geopolitical trap that often snares these bigger deals. The more popular outsourcing destinations are among the fastest-growing emerging economies. What we knew about them yesterday--cost, labor market, political risks, legal system--will likely change tomorrow. The more work placed in one locale, the more uncertain the outcome.
In 1952 Colin Chapman founded Lotus Cars with the intent of building small, lightweight sports cars that would outperform competitors that had been burdened by complexity. The company was the first to win 50 Formula 1 world championships, and produced a car with a design that went unchanged for four years, winning two consecutive championships and still winning races in its fourth year. Chapman was often quoted as saying the key to winning was to "simplify, then add lightness." Time has proved him right.
Far too often outsourcing creates new bureaucracy that aims to help smooth buyer-vendor integration but ends up isolating client stakeholders from vendor teams. This creates expensive delays, broken dependencies, unnecessary work or unused products. Rather than inventing an outsourcing-specific set of practices (and practitioners), we should borrow from those who came before us. For example, lean manufacturing (and its predecessor, the Toyota Production System) does an excellent job in predicting process inefficiencies.
In software development simple things work: Assign every project a business owner, give engineers direct access to stakeholders and encourage teams to experiment, fail early, and learn. Outsourcing professionals today are well-advised to think less about what makes outsourcing special and more about what has been proven to work.”
Alexei Milleris a partner at DataArt, a high-end software outsourcing firm with industry-specific expertise in the financial, travel, mobile and health care sectors.