Arbitraging costs and knowledge on the periphery: the embeddedness of software development outsourcing to the Ukraine

1. Arbitraging costs and knowledge on the periphery: the embeddedness of software development outsourcing to the Ukraine Jane Hardy Business School, University of Hertfordshire

2. THE CONTEXT Four new developments in global economy; • Shift to foreign investment in services • Horizontal to vertical investments • Offshoring skilled work • New host destinations for offshoring

3. Empirical focus: Offshoring to the Ukraine • Highly qualified students in maths and theoretical physics • Population 49 million • Key cities Kiev, Lvov, Kharkov and Odessa • Low levels of FDI • ‘Peripheral’

4. INTERVIEWS CONDUCTED

5. THE DEBATE Benign properties of ‘knowledge economy’ assumed rather than interrogated Scenario 1 • Possibility of regional growth and dynamism from capturing high level service functions? Scenario 2 • Fleeting comparative advantage as low labour costs for skilled workers provide a brief window for arbitrage?

6. CONCEPTUAL FRAMEWORK Three influences on embeddedness adapted from Zukin and DiMaggio, 1990; White, 2004. • Structural (features of sector) • Cultural and cognitive (Leveraging knowledge) • Network (Role in value chain or corporate network)

7. STRUCTURAL INFLUENCES ON EMBEDDEDNESS Definition ‘…the parameters that limit the field of action in which agents formulate strategy, and broad imperatives which ultimately push firms towards particular ends, albeit via a number of diverse routes and managerial strategies (SCHOENBERGER, 1994).’

8. Dimension 1: Cost and knowledge seeking By leveraging our global facilities, world-class engineers and state-of-the-art tools and processes, companies can increase the quality of their software products while dramatically reducing timelines and operating costs - often by more than 60% (Senior Manager Ebuilders).

9. Different routes to arrival - Path dependent - connected to previous personal & business connection (subjective) - Matrix comparison (objective)

10. 0= very little/none 1 = enough 2 = good 3 = very good

11. Dimension 2: Spectrum of skills If you take one extreme as commercial websites where there is very little algorithmic content and computational work, it involves putting up screens, moving data from here to there and putting it into a data base. But also high skill: They [Ukrainians] had a lot of expertise, putting it bluntly, in calculating the trajectory of a missile…so very strong depth of talent…not a strong local market (Marketing Director Globalogic)

12. Dimension 3: Hypermobility • Sunk costs only in training • Relocation in a series of waves I’ve got a friend his model…is he goes round sets up and office and hires people…and he sells the bodies.

13. Every country in the world is a possible site for software development. If one country is too expensive you just go to another country …There are computers and intelligence all over the world..it is easy to find out what they are doing by the code they write..if they don’t write good code you get rid of them (CEO Extremebyte)

14. COGNITIVE AND CULTURAL EMBEDDEDNESS Cognitive and cultural embeddedness refers to the individual corporate strategies used by firms to leverage knowledge across national boundaries

16. TWO MODELS FOR ACCESSING KNOWLEDGE

17. Logiglob & Heavensent: relational model There is a traditional way of building systems, which is known as waterfall where you have an idea and write a spec, you take that spec as though its gospel, develop the product and then it often has no value to anyone as things have moved on. There’s a very much more agile approach where you’re my business user, you’re continually coming up with new ideas, so I am building in short iterations, sometimes as short as two or three weeks. Everything is evolving – you are very rapidly going through that life cycle.

18. Value setting Value setting is important.. we try and have one company offsite once a year in Seattle or in the Ukraine..somewhere where you all get together and the first thing you do is talk about values and shared values and you go through a value setting process to define company values. We have these marvellous cards and basically people move cards around until everyone can agree as a group.. it works great.

19. Commitment engendering ‘ Body shopping is all about you want ten people, …interview some or all of them, that’s what you get, they’ll do what you tell them to do – that requires much more structure and discipline on your behalf to be successful, we are far more tell us what you want at a relatively high level and we will help you with the whole process, but what you really want is not fifteen engineers, you actually want a product out by 31 March….so we can focus on the end deliverable (Logiglob).

20. Employee engagement The technology makes it possible to have a cultural D&A across the organisation, you couldn’t do it without the technology, but technology doesn’t make you creative, technology enables you to act in a certain way, it doesn’t turn a hierarchical organisation into a flat one, a moving agile organisation, but it does put the tools in place to allow it over long distances, without the technology the only way you could function would be hierarchically, because you wouldn’t have the medium for feedback and interaction. But just putting in that medium doesn’t make it work.

21. Employee engagement • The daily stand up is a quick group by group where are we at, but importantly what they are working on and what problems are they facing. Hearing that someone is having a problem in one area quite often promotes someone to say I can help you with that.. so it gets that flow going. (Cloudmade)

22. Extrembyte: Non-relational Hierarchical and modular These people are here to get a job done, they don’t care what we look like and we don’t care what they look like…We track them by the numbers, we estimate in advance how long each task should take, and if a programmer is consistently taking longer to do a task, then we let them go.

23. Modular task accounting …people can walk in…walk out, see what needs to be done…get the code…download it, work on it and upload it again…it can be done anywhere…they could be on a beach in Tahiti if they wanted to be. The entire company is virtual there is no company in the US either. Our costs of doing business are as minimal as possible

24. Non-engagement with employees • Communication ‘gratuitous’ • Meetings ‘a waste of time’ So many managers, so much design, there would be meetings, everyone would coming for the meetings, then there would be pizza…there would have thirty people in a meeting, but by the time the code came out it was piddly it was nothing…it was just really inefficient.

25. NETWORK EMBEDDEDNESS The role of the activity in the corporate structure or value chain in respect of the level of skill and/or degree of control and autonomy.

26. 2. Unstable networks • Financed by venture capital • Continual cycle of dissolving companies and the opening up of new ones, often by same managers.

27. 2. Centrality of teams & projects • All firms are building, buying or selling teams We don’t build any of our products and we don’t have any of our own intellectual property…You come to me and say Dave – ‘I’ve got this software company and this is the product I’d like to build I would then assemble a team for you and that becomes your virtual subsidiary.

28. Technically we only have one employee, all the rest are subcontractors and self employed and responsible for paying their own taxes...We have seen the emergence of people whose firm is really just finding bodies and truly just giving you a body, so that you can integrate them into your team.

29. 3. Kaleidoscope configurations • Sector is much more turbulent in terms of the life span of companies • characterised continual cycle of dissolving companies and new firms opened by the same managers, mergers and acquisitions and takeovers. • Kaleidoscope, where the industrial scene is constantly morphing into complex new configurations. • Ukrainian employees spun off from firm to form own company, who then became sub-contractors (and absorbed risk)

30. TENTATIVE CONCLUSIONS Positive • Presence in 4 cities • Employment for highly trained graduates • Transfer of expertise in management • Opportunity for spin offs • Locked into global networks with customers

31. Negative • Weak territorial embeddedness & few multiplier effects • Wage competitiveness only temporary • Turbulence • No intellectual property rights • Small segment of value chain • Spectrum of employment security