I have just discussed that the overall best way to go with your Big Data project is to start with people inside your company. However, there will be times when you will need assistance from outside vendors. It is important to show the same care and the attention you spent on choosing your Big Data team in choosing your Big Data vendor.
When working with third parties, it is important that you treat them as a temporary support. I mean, when you are receiving outside support, you cannot let them do everything for you, and you having no clue as to what they are doing. Your responsibility is to monitor them carefully, learn how they are carrying out tasks, how they are performing their jobs and to document everything. Big Data is no different.
Although I have expressed my ideas with reasons against Big Data outsourcing, there will be times when companies will need to outsource their Big Data projects. The reasons may be plenty, such as infrastructure, required investment, ROI concerns, staffing issues, budgets or some combination of these.
If Big Data is to be outsourced, I advise you to start with a pilot project. Big Data is a new area in IT and so are the contractors. The contractors will try to reap profits from the hype and will be paving ways for a long term engagement to ensure that when the hype is cooled, they will still be making money. In order to protect your interests, ask for a pilot project to evaluate the contractor. And watch out for the signs to tie you to a long term engagement. It is you, not the contractor who will decide if the contractor is well worth for a long term contract.
The contractor needs to understand your industry in detail rather than having a vague idea or with the unjustified self-confidence that his expertise in one industry will apply to yours as well. Remember that your industry is different and your company is unique and your contractor’s cluelessness is his own problem. The contractor must have the skilled personnel, the right tools and the right processes to work closely with your company. If you already have a Big Data team working and you are hiring a contractor, make sure that the contractor can deliver what your team asks. If you don’t have the dedicated Big Data team, make sure that the contractor can effectively communicate with your business-side (non-IT) personnel and deliver their requirements (rather than steering/convincing them that what he delivers is better than what they ask).
In parallel to all of these, remember that what you are outsourcing is your data, which is your company’s most valuable asset. Make sure that the contractor/vendor is tightly bound by the contract terms from legal, financial and security points of view. You will need a very detailed non-disclosure agreement because it is the key of your company that you are outsourcing. To have a better peace of mind, engage your IT security and company lawyers in the project, put control points and keep the executive management very closely informed on the matters before the data leaves your company.
One of the most neglected points in the consultancy/vendor contracts is the documentation and the training issues. Almost all of the companies act like outsourcing is a one-way relationship: when the contractor/vendor is decided, everything is their problem. But it is just the opposite. When you are giving things away, you are giving your knowledge away. If, at some point in time, you decide to take this knowledge back, and if you did not receive anything from the contractor during the period, you are left with a ticking bomb in your hands: a machine that is working, which you don’t have any clue about. To avoid such a situation, make sure that what the contractor does is properly documented and his knowledge is returned back to your company. If you have an in-house Big Data team, this knowledge will in turn contribute to your team’s level of knowledge and expertise.
If you are hiring a contractor for training purposes, there are a couple of points to check to make sure that the training will be effective:
Does the trainer have a hands-on expertise with Big Data so that he can answer your team’s scenario-based questions?
Does the training facility have the infrastructure for a hands-on Big Data training?
Does the training facility provide training for one vendor only? (This may mean that they have a rather limited – vendor-specific – perspective on Big Data.)
Does the training offered addresses the needs of your Big Data team’s requirements? (Did your Big Data team reviewed the training outline/contents?)
If you are ready to outsource your Big Data to a third party, rethink once again. How will you make sure that your Big Data project is not tied to the success of an outside organization after you have given them your company’s most valuable asset?
Featured Image: www.greenbookblog.org