Do a pilot not a demo, when selecting outsourcing partner
Published by Guntis Urtans on Friday, December 07, 2012 16:35
Frequently our customers who have software development work to be outsourced, decide to start with a pilot. But beware of the exact definition of the pilot. In selling mode, some vendors execute something completely different to leverage the sales opportunity and you might end up with a big surprise.
The pilot approach makes sense, sinceyour position to discover shortcomings, issues and in-efficiencies with the vendor is suited to small, usually not business critical projects. If the timing is not critical for the larger project, this is a good way to start.
On the outsourcers' side the suggestion to start with a pilot project is frequently used. Usually it comes with: "try a small pilot with us and you will see how good we are".
My advice to customers is: Yes, it is good to start with a pilot if you lack deep experience in outsourcing, or for a specific type of projects.
Pilot or demo
But - do not do the pilot just to check the vendor's capabilities. Why? - you may get a demo instead of a pilot. A pilot should bring you working and tested, production grade software that brings value to you. A demo is something else.
The demo is usually impressive, since it is done for a specific purpose - namely to demonstrate that it works and that the vendor is capable of delivering it. But there is a significant risk that you will have a completely different picture of it, when you need to scale, the complexity increases and you are under time pressure.
So - do a pilot as a proof of concept, not for validating the vendor's capability.
Things you may find useful:
- Use a pilot as an opportunity to set-up and validate processes. Outsourcing usually affects processes on the customer side. This is an opportunity to formalise changes, get feedback from your organization and fine-tune it.
- Validate your assumptions about the outsourcing scope. A pilot project can potentially show that the vendor is capable of doing more or indicate that some of the activities you were planning to outsource does not work as expected - so it would be good idea to postpone it, or give it a second thought.
- Set your expectations and performance gap analysis during the pilot project, as well as after the completion of it. Use the pilot to tune your metrics.
- If the pilot fails, do a root-cause analysis. We are frequently facing statements "we tried it once, it did not work". Very rarely is the concept of outsourcing entierly wrong, but it is hard to make it work for you.
There is no logical reason why outsourcing of software development shouldn't work, unless it is a core activity and your competetiveness depends on the software develeopment skills of your trade secrets. Even in those cases we see an increase of outsourcing activity.
There is a zone where outsourcing of software development is recomended, and you should try by using a pilot project. If you make sure you are really getting a pilot and not a demo and it still fails, it is more likely that you need to make some adjustments to your own approach and possibly be better prepared for it internally.
Some reasons for the failure can be due to your vendor selection, such as lack of technical or domain skills and cultural differences: Others can be on your side with wrong expectations, lack of comparative metrics and business alignment, lack of motivation from your stakeholders and teams, etc.
It is important to understand this and to do your homework - you will get it right the next time.