As retail customers, we have been probably exposed to the automated, technology oriented retail banking side of banking; where the operations management approach is applied to the branch('factory') to maximize flow('loans processed, accounts opened etc). Data mining is used to generate insights and make those offers of 'preapproved loans', 'life time free credit cards'. 'boosting credit limit' etc. Little personalized human insight/attention is given in those decisions.
But on the wholesale side, it all boils down to relationships. Given that the platform('infrastructure') of banks is quite similar nowadays (or can be outsourced easily), the dividing factor boils down to trust and relationships. Even where the balance sheet/financial projections etc may not support a positive credit decision, the relationship may tip the balance in the favour of approval. And given that there is no uniform price/price list for a transaction(there is often a deal RORWA though), there is ample discretion on the client facing executive to structure the deal. Agency problems are mitigated by P&L based bonuses, offloading the risk(thus forcing pricing to be realistic to avoid facign a loss) etc.
So which approach is better? I guess where someone needs credit(in whatever form-prepaid swaps, loans, guarantees etc), they are better off with a personalized approach where someone counsels them on presenting the case better, acts in their interest etc. But for transactional business, relationship managers may just be an unnecessary overhead unless they can leverage their client business understanding to add value.
As a small guy entering the banking door for the first time, I'm probably better off with a relationship manager, than navigating a morass of weird technology platforms. But the reality is that I will not get one, till I'm too rich to need them!! Financial literacy campaigns and outreach centres do help bridge the gap, but they seem a case of too little too late. So in balance, I would support a banking business model which uses technology to build a LOW cost platform, but which vests discretionary power(within limits) in the front office to offer better terms to the client. This may result in some corruption, but overall the financial inclusion objective would be served better.