When to Outsource

Knowing how to outsource a process is challenging enough, and requires serious operational management and help, but does not involve making strategic decisions.

Conversely, knowing when to outsource is far more challenging, as it involves making decisions with imperfect information about the future.

Caveat: Use this as a starting point, a framework, but do not use it as your sole decision-making process. Get serious help; we are here.

The Why

There are only two reasons why you should outsource something. In truth, they are two sides of the same coin:

  1. The outsourcer can get you better business results.
  2. The internal process has negative impact on the rest of the company.

Better Results

Usually, you already are thinking about outsourcing because it gives you better business results. Most of the time, it is lower costfaster delivery, or higher quality. Of course, these are the same thing. If you can manufacture twenty books an hour, and an outsourcer can manufacture forty books an hour for the same price, they aren't really faster. You, too, could get to forty books an hour... but it will cost you more to retool and get there.

What you really are paying for isn't twice the output; it is twice the output at the same cost. Or, if you prefer, a lower cost per book per hour.

Similarly, it may take you two weeks to order a new HP or Dell server, rack and power it, install a hypervisor, and be ready to install applications. It will take you minutes on Amazon Web Services. Could you do it that quickly in your data centre? Of course you could! But it requires more spare capacity and power all along, and extra software to manage it, all of which will cost you more.

In the end, we could more accurately call this "Lower cost per unit at the same quality level and speed of manufacturing," but "Better Results" is much more memorable.

If your reason for thinking about outsourcing is "Better Results," be absolutely positive that you know exactly what results you expect to get, that you accurately measure them now, and you know how to measure them accurately afterwards.

Perhaps most importantly, be honest about the metrics. Executives rarely lie or obfuscate, but biases subconsciously insert themselves into everything we do. Someone with a vested interest in the already done outsourcing is unlikely to catch a failure early. If you can, have someone other than the executive who owned the outsource own the metrics.

Even better, be sure not to compensate the executive on the basis of outsourcing success or failure. Every business process change is an experiment. You need to be comfortable enough learning from both successes and failures and reversing course when the situation warrants.

Negative External Impact

Sometimes, you are not outsourcing because of the direct outcomes of the process - cost per unit, speed, quality. Instead, the internal process is causing significant distraction to other parts of the business and/or executives.

This is a much more amorphous reason to outsource. If a process has issues, and is eating up the COO's time, for example, then something is broken with it. As we said previously, outsourcing will not solve your problems. Most of the time, outsourcing a process because it has a negative impact on some other part of your business is a really bad reason to outsource.

Instead, you should fix the process internally, and then use standard business metrics, from above, to measure how well you can do it internally vs. outsourced.

Nonetheless, every business process in a company requires some executive time and input, and has some impact on other parts of the organization. If a process exists that really does eat up too much of an executive's or other group's time and/or has a negative impact, and you already have fixed it internally as much as possible, then:

  1. Measure what the current impact is in standard business metrics.
  2. Measure what the outsourced impact would be.

Does this mean that "Negative Impact"really is just a variant of "Better Results"? Yes, it does.

If you cannot measure it, you cannot outsource it.

So if all of the criteria are met? Should you do it? Not necessarily.

Even when all of the ducks apparently are lined up, there are counter-balancing forces to take into account. Our next outsourcing article will look at some of those.


Once you (a) know how to outsource, and (b) are positive a process is ready for outsourcing, you have to be able to prove a process is a good candidate for outsourcing. It always comes down to quantifiable business metrics. If you cannot measure them, outsourcing will fail, without your even knowing it. Ask us to help you measure them properly.

It is important to measure your executives in a way that allows them to experiment with outsourcing when it makes sense, without being penalized or creating biases that prevent further flexibility.

Finally, know when not to outsource...