A stalled outsourcing decision

Pragmatic criteria and techniques businesses can use to reduce the risk of a bad decision.

Dear Bob -

Our company is in the process of deciding if we will outsource a core part of our business. The stakes are high in this decision because of the potential financial impact of a wrong decision and the potential impact of a degradation of service.

What makes this decision difficult are the unknowns. While we can check references and even define service terms in a contract, we cannot be sure that we will get the same level of service that we are used to. From a financial aspect, there are many different measurements that we can define historically from our data, but we cannot get a solid feel for with the vendors we are talking to. This is due to the nature of the business and the confidentiality of the vendors business processes/methods.

The final struggle are the obvious complications of outsourcing and integration of systems. Not only are the processes involved, but there are several options to the final implementation and an ala carte menu of options to outsource.

We have some internal leaders who believe there are savings to be had by outsourcing, yet we cannot develop the proper evidence to make a sound decision.

Do you have any ideas on how to better approach this decision?

- Dazed and Confused

Dear D and C ...

I think we need to approach this from three directions to get to a satisfactory resolution. The angles that occur to me are the nature of evidence, risk management, and alignment of purpose.

Evidence first: You've gathered what evidence there is, and enough uncertainty remains that this isn't a sure thing.

Depending on the nature of the business process you're thinking of outsourcing, you might be able to gather better evidence. Many of the outsourceable business processes are transaction based or case based. If that's true here, pay the outsourcer to process a few months worth of transactions or cases, just as if they were real, and provide you with the results for analysis.

A comparison between their processing and your own would, I imagine, reduce the decision-makers' uncertainty to levels low enough to support a decision.

I'd expect the outsourcer would be willing to charge a discounted rate for this service, accounting for the balance as cost of sales.

Risk management: As a side-note, different types of company are better at this than others, and it works the opposite of how you'd expect it to work.

Insurance companies are worse at this than, for example, entrepreneurships. I think it's because insurance companies operate in statistical universes that more or less eliminate risk: If you know the odds for any individual policy you can calculate the rates you have to charge to cover the expected number and size of claims. That's what actuaries do.

Entrepreneurships are more accustomed to dealing with situations that are statistical samples of one. It's more meteorology than underwriting -- more "60% chance of rain today" than "2% claim rate with an average claim of $35,000."

You can deal with risk in three ways -- reduce the odds, reduce the damage, or buy insurance. In this case, the purpose of due diligence is to reduce the odds. Unless you're able to buy the analysis I described earlier, you've pushed this to its limit.

You can buy insurance, in the sense that you can write penalties into the contract for failures to meet defined service levels, but there are serious limits to the value of these. If you've bet the farm on the outsource, your own claims process infrastructure will be gone, after all.

That leaves reducing the damage.

If it's possible given the nature of the outsource: Establish a progressive roll-out, with clear criteria for making go/no-go decisions after each new segment of the business has been handed to the outsourcer. That reduces the damage in two ways. First, you aren't betting the farm all at once; second, you don't eliminate your in-house capability until you're sure it's going to work.

If this was a systems conversion we'd be calling this an incremental implementation with a roll-back plan. Same principle. Outsource in stages and make sure you can undo the outsource if it isn't a clear success.

Alignment of purpose: It's worth asking if everyone agrees as to the point of all this -- whether it's to be less filling or to taste great.

There are six possible goals for this sort of thing. They aren't mutually exclusive. They also aren't mutually reinforcing. They're the six optimization parameters for any business process: Overhead cost, unit cost, cycle time, throughput, quality (absence of defects; adherence to specifications), and excellence (flexibility, ability to customize or add unique, value-adding features to products and services).

The main benefits a company gets from outsourcing are usually:

  • Reduced overhead: You get to dismantle infrastructure, transferring the overhead to the outsourcer.
  • Reduced transaction (unit) cost: The outsourcer has economies of scale you lack.
  • Improved quality: Because the outsourcer operates according to defined rules with a much reduced ability to apply discretion, nuance and judgment to situations, fewer claims will be out of tolerance.
The primary process tradeoff is reduced excellence: It's the flip side of improved quality. The flexibility of the process will be significantly reduced. The primary business trade-off is, as already mentioned, the risk that the outsourcer won't deliver.

If you haven't already done so, make sure everyone involved in the decision agrees as to what you're looking for: Reduced overhead, reduced transaction costs, or improved quality. ("We want all three." "That's fine, how do you rank them in order of importance?")

It's entirely possible at least one decision-maker has a different goal entirely -- the desire to make this business function Someone Else's Problem. It wouldn't be the first time a business outsources something for the wrong reason, and this is definitely the wrong reason. Managing an outsourcer is no easier than managing an internal executive. Far from it.

Making sure everyone has the same goals is important for when the most likely outcome happens -- results that aren't good enough that the outsource is considered a spectacular success, but good enough that you can't easily decide it's a failure.

One last point: I'd strongly suggest your CEO establish a clear rule in advance, which is that a tie goes to internal processing. If it turns out the financial benefit is neutral and quality is as good but not significantly better, then you cancel the contract in order to eliminate the future risk of the outsourcing relationship going sour.

Big decisions are hard decisions. Big consensus decisions are even harder. I hope this helps guide your company's executives through it.

- Bob

Copyright © 2008 IDG Communications, Inc.

How to choose a low-code development platform