This is post was translated from Dutch using a locally run copy of the Mixtral LLM, using a very low-end GPU. I did some light editing here and there, but the translation is mostly fully automatic.

The short version: organizations often hesitate for many years before outsourcing tasks, particularly in the field of ICT (Information and Communication Technology). During those years, valuable ICT employees leave because constantly justifying their own existence is frustrating. In the long run, something will eventually go wrong, making the decision to outsource much easier: “We can’t and don’t want to do this ourselves anymore.” However, by hesitating for so long, you have created the situation yourself. This article contains a list of criteria to help you decide why you might want to keep certain tasks in-house. Use such a list to make a quick, solid decision: what to outsource and what not to. Then, maintain & value a smaller core team of ICT professionals to handle your own tasks well and assist in keeping your outsourcing partners on their toes. This way, you prevent ever having to outsource the heart of your organization, which rarely turns out well. Above all, make these decisions consciously and quickly, and avoid creating a years-long dramatic situation.

The organization responsible for managing “.NL” recently made headlines for their decision to outsource themselves to a Canadian and American company. With an annual revenue of 20 million euros, they used the argument that they are too small-scale to be able to take responsibility for their own technology. The Dutch Parliament also uses this argument to place their email under American management.

Announcement of the outsourcing of all servers to Amazon. Source: Linkedin

This discussion is taking place in many forums. Companies and organizations frequently have to make a choice: do we do it ourselves, or do we outsource it? In the past 25 years, I have seen this play out many times and have spoken with numerous technicians and management professionals about it. That is why I think I can still contribute another useful piece on this subject.

So, this has been going on for quite some time now. Many companies have first outsourced their customer service as much as possible, only to discover years later that it is fundamentally very difficult to help customers from the outside. The problems mainly lie within and can only be solved from within. By now, a middle ground has been found in many cases - helpdesk staff are still temporary workers, but “support” is once again part of the parent company. With AI, we will likely have a new round of experimentation in customer support.

In the past, companies had “coffee ladies”, mechanics, chefs, cleaners, and even their own printers on staff. In 1999, Casema, a precursor to Ziggo (a large cable company), still owned its own printing press. These are things that have now been outsourced everywhere.

At the moment, ICT is on shaky ground in many places. Do we do it ourselves or do we consider it like plumbing: we call someone when it stinks. There is something to be said for both approaches. If you do it yourself, you have a lot more control, but you also get dirty hands. And if you don’t do it yourself, you just have to believe what your supplier tells you. Because if you no longer have expertise in-house, it is difficult to know if your ICT partner is doing a good job. And without actually still running some technology yourself, it is also difficult to retain expertise (more on this later).

In considering whether to handle something in-house, three factors should be weighed:

• How important is this topic for my company now?
• Is it something that determines my future, can I still innovate?
• How difficult is it for me to do it well myself?

For example, a large law firm with sensitive clients. A law firm stands or falls by its confidentiality. Communication must therefore be exceptionally well protected. But, email is ultimately not what the firm does. The product is legal advice, not an email service. However, if IT offers better possibilities for collaborating on documents with clients, that does provide a competitive advantage. Furthermore, doing good email yourself is challenging and requires a dedicated team of system administrators who are hard to find.

This last factor means that only very large firms can still manage their own email or even run a platform for working together on documents. For medium-sized or smaller firms, it is simply not feasible at this time, so they should look for the best partner (which is not always easy to find). In practice, no one does it themselves anymore, which means that a law firm wishing to sue Google often has to do so via servers managed by Google.

Another example is catering. It is important that people who come to the office are well fed and hydrated. But, it is not something where companies can really distinguish themselves. A very good lunch provides only limited benefits (for example, in recruitment). Meanwhile, professional food preparation is surrounded by mountains of rules, certifications, inspections, and hefty fines if you do not comply. All three criteria therefore point to ‘better outsource it’ for all but the very smallest companies. An exception are restaurants, whose role it is to prepare food. And that brings us to the situation that is currently in the news.

SIDN manages all internet names that end in .NL. Properly managing these 6 million names is the lifeblood of the company. Technically keeping the .NL names in the air is the core competency of the organization, and also determines its future. If they don’t keep up with the times, they may lose the right to exploit .NL. Furthermore, it takes top people to keep a “country code domain” running smoothly. But, since this is also the core task of the company, it should be possible to train or find these people. For techies, it is a great honor to be allowed to do such a thing.

In short, all three criteria point to continuing to do it yourself. If .NL is not capable of managing .NL, that is like a restaurant buying its meals from a delivery service.

For some organizations, specifically government services or national registers (such as .NL), there is an additional consideration. As a country (or Europe), wouldn’t you want to be sure that you can conduct your own (government) communication and elections, even without the help of people on two other continents? A report commissioned by the Dutch intelligence & security service AIVD, made by think tank Clingendael, was published today and it discusses this situation.

The decision to do something yourself or not

Doing something yourself or not is a very significant choice, and organizations often waver over this choice for years. They give up small things where the choice does not seem to have a lot of impact. But they then struggle with big things without investing in them any longer. There are still in-house professionals with relevant skills, but management & shareholders openly discuss letting go of these people.

This greatly undermines the morale of these professionals, who must constantly fight for their own right to exist. The (good) people who can easily leave do so in the meantime, because there is nothing fun about working for an employer who would rather see you go. It is also challenging to recruit new people, as applicants quickly sense this. And current employees also advise against coming to work there.

Subsequently, things sometimes go wrong, not helped by the disrupted labor relations, after which the choice is often automatically made: we can’t and don’t want to do this anymore, we will outsource it.

But this was not a free choice; one has allowed it to come this far. You make good choices yourself, you don’t let them happen as a result of self-organized incidents.

A feasible solution

There are topics that are not very relevant for the mission or success of the organization and for which there is also plan B possible. Compare this to catering in the canteen. You do not have to know anything about HACCP to arrange your catering well. If you can’t figure it out, everyone just arranges their own lunch or has it delivered for a while. It is not ‘mission critical’ in the short term.

For IT, there is a tougher consideration. This is often really relevant for the mission and the future. Nevertheless, there are also parts of IT that you can do without (payroll software, sending newsletters, etc.). What remains is a core that you explicitly have to account for as your own business. For the .NL people, this includes managing who owns which domain name, customer contact, and actually keeping .NL in the air.

The essence is to clearly establish early on which systems you want to retain. Previously, your IT department did 20 things, and now only 3. However, because this core group of services is so important, you should take your remaining IT talent very seriously. They are, after all, responsible for the core of what you do. And by giving this smaller group of people an important place in the organization, they will also take good care of the 17 services that have since been outsourced. Because unlike the cafeteria, you need to be a well-informed customer of your IT suppliers. Otherwise, things will still go wrong, or you will at least end up with enormous invoices.

An excellent situation is therefore a small IT department that keeps a few very important things in the air and thus remains hands-on and knowledgeable about how things work. This also attracts good people. And in the meantime, this small IT club can effectively manage and control your other suppliers.

Your remaining ICT employees will then change from being a burden to valuable success factors for your business, involved in all important decisions.

What usually happens

The above may be ideal, but it is difficult to achieve. The transition from “we are trying to let go of our IT employees” to “our remaining IT employees are worth their weight in gold” is almost impossible. Because all the good people have already left, and we can no longer form an A-team. Additionally, after years of fighting over doing things ourselves or outsourcing, relationships have become severely distorted.

This is partly due to the fact that external suppliers generally do not try to sell their services through good ICT employees. Instead, they work on the management by salespeople with beautiful PowerPoints, who thus polarize the higher management of the organization against their own computer talent.

There’s no need to do everything yourself by the way. Source: Wikimedia

You end up in a situation where everyone who loves IT has left, and as an employer you have also become very unattractive to people who have actual skills.

What to do now

Many organizations now find themselves at the end of this battlefield, and have outsourced too much. I have written about this situation earlier on several occasions.

In short, you have to transform your organization into a “master outsourcer”, which in itself is possible, but you have to fully recognize your situation. A good example is Odido (formerly T-Mobile NL). This telco is almost completely outsourced. Recently they changed their name and the management gave a week-long tour of how important this rebranding was for them and how they managed to pull it off. But there was no word about the actual telecommunications technology that this communication company provides. It’s not that important to them. Odido is a kind of ‘hub’ through which all external providers sell their services. Even most antenna masts are no longer owned by Odido, for example.

Personally, I would not like to organize my business in this way, but if you do, do it like Odido. Embrace outsourcing and moving services with conviction and then focus on marketing and sales. And incidentally hope that your suppliers will continue to agree with your future vision (🤞). By the way, being a master outsourcer still requires a lot of qualified personnel, and it is quite a struggle to find and retain them. Because your techies need to be smart, but can’t sit at the controls for the most part.

If your organization is still in the outsourcing process, you have more options. “Soft healers make stinking wounds”, so tough decisions are welcome. But also know that it is incredibly difficult. The push to outsource everything (to “the cloud” or to concrete parties) is enormous. Almost irresistible. Parties that should really know better run tenders to get even the most crucial things out of the door.

Concretely, as recently discovered by Joost Schellevis, both the Second and First Chambers of Dutch parliament have now handed over their emails to Microsoft. They thus make policy about Microsoft via Microsoft servers, and communicate about our international relations through servers that are directly accessible to the US government . Good luck with that!

Apparently, under this kind of pressure, it is difficult to make rational decisions, but we will have to try.

In Summary

Choose quickly, using good criteria (see above for inspiration), the services that are truly crucial for the goals and future of your organization, and explicitly gather the best people for these to do yourself. This may require significantly higher salaries and a much better position and appreciation for your remaining IT personnel. There should be no further discussion about the selected services: they are ours. And push the rest away as quickly as possible, led by those who are certain that their role is not up for discussion. They can then help you keep your suppliers sharp.

By explicitly and quickly making the “in-house/not in-house” choice, it is still possible to attract the right people on time and make a success of your transition.

And this way, you prevent having to tell your customers one day that despite a turnover of 20 million euros, you are no longer able to perform your core responsibilities yourself.