On National and Enterprise Outsourcing

In the past, companies and countries were extremely self-sufficient. Doing business internationally was cumbersome, and the market for business to business services had not yet developed to the scale we know today.

In this post I explore how companies evolve as they outsource themselves, plus how countries do the same thing, and what the consequences are. Finally, I discuss how hard it is to regain abilities that have been relegated to suppliers.

Peripheral outsourcing

If we focus on companies first, functions that are now almost always outsourced were previously almost exclusively performed in-house: payroll, cleaning, office/building maintenance, catering, internal mail delivery, marketing and even printing customer communications at scale. Over time it became clear that employing cleaners is very far removed from what most companies do, so it makes sense to have that market concentrate, and instead contract office cleaning services. Employing one thousand cleaners is way more efficient than having to manage 4 of them.

Similarly, doing payroll is complex and the consequences of messing it up are huge, so it is more efficient to get a third party to do this. Balance sheet mechanics as well as a more highly developed real estate ecosystem meant that most companies also got out of the business of building, maintaining or buying offices.

Over time, more and more peripheral functions thus left companies themselves - and this is not immediately a bad thing. Excelling in cleaning offices does nothing for customers or the bottom line, unless you are a cleaning company. If an itnernet service provider does its own payrolling really really well, this does not lead to a competitive advantage.

The upshot is that (for example) a modern marketing company might literally only have a marketing staff - office, cleaning, employee administration / HR, IT are all outsourced and/or in the cloud. This has resulted in a far more dynamic marketplace of providers - previously “launching a company” required acquiring large amounts of capital goods & staff to administer these before the actual work could begin.

Even this peripheral outsourcing however comes with a downside. In the past, companies had to do so many things just to be viable that they had to maintain a general level of competence at doing all kinds of things. Such companies were more able to things outside of their immediate core business because sufficient people were around with diverse skills. Even large companies were able to get stuff done (from time to time).

A modern large enterprise is often literally unable to organize a customer event (‘do we expense all this? Is the local supermarket sufficiently certified? Do we have to run an RFP?’) and then ends up hiring an external vendor to do it for them - at great cost.

In addition, as a company becomes less adept at ‘doing’ things, the remaining staff will focus more and more on contracts and the legalities of outsourcing, spending a lot of time on certification and compliance. This makes it ever harder to do anything novel - “is it safe?” “is this startup ISO 27001 compliant?“. And this in itself makes outsourcing ever more attractive - it is too much of a hassle to actually do anything.

But why stop at the periphery?

What is peripheral, what is core? Why only outsource the cleaning staff? My home turf consists of very large scale telecommunication companies. In such places, often national incumbents, over the years most ‘core’ activities have also been outsourced. This may start with the process of digging cables, which could indeed be said to be somewhat peripheral - lots of places can do that for you, it may make no sense to have a ton of equipment on hand even when no new cables are being laid.

It may also be the case that a large scale rollout of a new networking technology is best done by a vendor - such large scale migrations (to 5G for example) only happen once every 5 or even ten years. A vendor will do such migrations much more frequently, and will be able to do it faster than the service provider itself. This too has some logic to it.

Outsourcing decisions are partially made based on cost, and an interesting dynamic is that when outsourcing, relatively speaking, the overhead costs do not go down as much as the actual operational costs. Or in other words, “HQ” is getting bigger relative to the rest of the company. This in turn makes every activity in the company more expensive, which makes the case for further outsourcing even stronger!

Thus next up follows the outsourcing of customer care, network maintenance, customer equipment logistics, radio network operations, IPTV streaming, recruiting and marketing. In this way, telecommunication service providers find themselves as husks of their former selves.

This situation is not unique to service providers - many other industries have also managed to “outsource themselves”. Examples include Boeing, as discussed in this seminal paper, but also much of the food industry. Pharmaceutical companies have similarly outsourced most innovation to startups, which they then acquire when the time is ripe.

How does this even work?

We may wonder how it is possible to stay successful as a company without having significant expertise left at being that company. I wonder about that a lot in fact.

The Boeing paper linked above ‘Out-sourced profits - the cornerstone of successful subcontracting’, written by visionary engineer L.J. Hart-Smith, spends a lot of time on the eventual effects of all this outsourcing. Some rather obvious questions are rarely addressed when doing large scale outsourcing.

For example, to find good companies to outsource to requires industry expertise. And once a partner is found, it needs to be supplied with specifications and instructions that make sense. This too is work for specialists, as is monitoring the quality of execution. The problem is that a company-that-does-not-do-anything is not a great place for a specialist to work. In addition, an enterprise that is very heavy on procedures and contracts may not even appreciate actual expertise anymore!

“We have a contract that says the partner will deliver”. Boeing has found out the hard way that a contract may say a lot of things, but it will not get your plane off the ground if the vendor doesn’t manage to get the job done.

Additionally, not only are the day-to-day things like procurement and vendor management hard to do without expertise, it is nigh impossibe to innovate and invent the future if the company is distanced from the actual industry - simply because no one knows what is possible outside of the pre-printed vendor brochures.

Eventually, heavily outsourced companies either regain their sense or they become so moribund that they get acquired, often by another heavily outsourced company, or perhaps by a corporation that is actually still capable and can outsource its new acquisition to itself.

The expertise is still there right?

If an industry has decided to move its core functions to third parties, there is an argument to be made that the expertise hasn’t vanished, it has simply moved on to other companies. This could then be construed as not being very bad in itself.

In general, and at first, this is of course true. Large scale outsoucing often in fact literally starts with transferring personnel from the customer to the supplier, often thousands of people at a time. The eventual goal is then often to hope that these expensive people go away so they can be replaced with automation or, more frequently, cheaper labour from abroad.

Hower, in terms on incentives something has changed in this process. Before the outsourcing, services were being provided to customers, and the people providing those services had a direct stake in their future success. In short, if they did better, the company would provide better and/or cheaper services, and there would be growth and rewards. Or conversely, if things went south, there would be direct consequences.

If multiple similar companies have outsourced parts of their core business to the same vendor (which is very common), there is no specific incentive anymore for the third party to do an especially good job. Just fullfill the contract and there might be a new contract. Excelling at the work will grow one of the customers of the outsourcing partner, but will likely shrink another. To a remarkable extent it has become a zero sum game. Just keeping the lights on is some sort of optimum.

I have a personal example of this happening - a major Dutch bank wanted to improve their DNS security, because belatedly they realized it might be a good idea. They had outsourced all this stuff however and the contract made no provision for DNSSSEC. The person in charge estimated it would cost around 200000 euros to get this simple change executed through their partner (for a single domain name). This is not the behaviour of a partner well incentivised to keep its customer successful.