Re: So, if outsourcing hasn't worked......
And furthermore Wipro, Tata et al are very viable alternatives if need be.
No they're not. ITOs share a common business model: Promise 20% savings against the incumbent provider, write a complex SLA where the vendor knows all the caveats and get outs (from experience) whereas the buyer knows none (doing this once every five years at most). Then take a seven year deal at a nominal loss, wait for inevitable variations and business change, and coin in during later years. There's even a term for this, "back loading", and my employers are currently in the phase that involves being reamed out by our IT "partner".
This situation is made worse by the common practice of building your ITO through acquisition, which leaves vendors with a balance sheet weighed down with vast amounts of "goodwill" (the amount they paid for acquisitions beyond their real worth). Unfortunately, the goodwill is capital on which the ITO have to make a return for their investors. This means that even if they do employ the cheapest of cheap, barely literate monkeys, the costs they have to recover from customers exceed the amount the customer was paying for inhouse and onshore skills in the first place, although because backloading gives a couple of years cheap it means that most customers can pretend that they've saved money. Although the converse is that those losses need to be made up by much higher charges in the later years, over and above the illusory "savings".
By the time the customer is paying the true cost, its been management musical chairs at the customer, and nobody remembers how much it cost in the first place, nor who made the decision to outsource. The skilled employees have retired, been P45'd or TUPE'd out, and nobody has the balls to even think about bringing IT truly back in house.
Outsourcing is the business equivalent of selling your kidneys - you only benefit on paper, and once you've done it you can't go back.