Why this might be the worst time to launch a consultancy

Officially, Defined Once will launch during Rustweek 2026. Though, last week we already introduced some nice new features on our website, including a git-backed diff renderer for our RFDs (which I thought was pretty cool) and our blogging setup. This got me excited to write a first post, sharing some early pre-launch feedback from clients and partners.

To kick off, Defined Once Engineering B.V., or just Defined Once is a new Rust consultancy company led by Max Gallup and me. It is "new" in the sense that the name and company structure are new, but Max and I have been working together for many years and many different clients now. Last year we realized that we could tackle more work by uniting our efforts, which also allowed us to think carefully about how we position ourselves. At the start of 2026, Defined Once was born.

Besides feedback on the name, we've received many questions about the course and the benefits of our offerings since then. Why Rust? Why consultancy? What about AI? Why now?

Calling it consultancy to avoid confusion

To me, 'being a consultant' has always carried a negative connotation because traditionally, consultants can be perceived as expensive and out of touch. They are happy to tell you what time it is, using your own watch and then bill you for it. The risk of becoming dependent on consultants or contractors is substantial: in the Netherlands for instance, recent reports have shown that consultants are starting to run the show at public bodies, slowly seeping away in-house knowledge.

When mixed with the worldwide geopolitical tensions, causing peace and economic prosperity to no longer be self-evident, saving on external hires is often considered an effective way to cut costs quickly. Add AI into the mix and the necessity for an external workforce is questioned even further. Could there ever be a worse time to launch a consultancy?

Of course, we are not deaf to criticism, and we acknowledge that the state of geopolitical and technological affairs doesn't make launching Defined Once easier per se. When Max and I discussed the content to go on our website, we thought long and hard on how to position ourselves differently: we know that we are not the "average" consultancy, but how can we show others? Maybe implementation partner sounds less out of touch, perhaps agency would better cover the operation. After some confusing meetings with potential clients, we slowly discovered why companies have dedicated teams for marketing and branding strategies...

In the end, we figured that consultancy is just the term that provides most clarity. Yes, we work for multiple clients and yes, we bill our hours per invoice. In that sense, we are just traditional consultants. On the other hand, we are also critical to big consultancies, expensive agencies or middlemen that add no value. We do not charge ridiculous amounts for questionable delivery, nor do we try to grow by hiring many interns and by placing them at our clients against crazy rates. We are a small collective of talented and passionate nerds - actually, at the minimal size to even call it a collective - that wants to focus on impactful projects with clients that we can build a sustainable relationship with. By uniting as Defined Once, we offer a broader field of expertise with better continuity than a freelancer could provide. At the same time, the commitment and risk for clients is far lower than hiring in-house, especially when you are unsure yet if Rust is the right fit for you. Our workforce can be easily scaled up our down.

While I think that our service is a great deal, I also recognize that's what most consultants would tell you. Besides, it's always tricky to add new people to your team, whether it's in-house hires or external consultants. Judging someone's capabilities, team spirit and work ethic is hard based on a PDF and a 30-minute interview. That's why I also want to distinguish our work based on a "30-day trial" period. After a month of working together we check in to see if it is a match for both parties. From there, we can make plans for bigger projects or longer-term commitments. Or, it could be the moment to part ways too, no strings attached.

On top of this, we are unique in that we operate in full transparency, with our founding values strongly in mind. This very blog post and its changes can be tracked through our public repository, and we try to give back to the community by working on open-source projects whenever possible. Not only because we think it's the right thing to do, but also because we have so much fun while doing so.

With this in mind, I am convinced we can distinguish ourselves from traditional consultancies. This is backed by our past and current clients, who appreciate our personal touch, or continuous effort to work together with their existing teams and our no-nonsense approach. We are not hidden away behind layers of management or "consultancy consultants". We're just engineers that are helping out, with the term "consultancy" just happening to create the least confusion about the company structure.

AI and software development

AI has, without a doubt, stirred up the world of software development. It has widened the skill gap, making the development of simple websites and web apps more accessible to those who have always had the desire and creativity, but lacked the technical background. Even for complex, low-level kernel work, recent AI models have shown very capable and useful. Some people have already declared the job of software developers obsolete. If not now, then surely in a few years. Not great when you are launching a software development consultancy.

Again, the observations are fair to a certain extent. The days of software engineers being forever irreplaceable are over. The idea of juniors being outperformed by agents is widespread and stubborn. It doesn't even have to be true, if the decision makers think that it's true, it can send shivers down the spines of many who thought that working in software meant never having to worry about job security at all. However, the first cracks are also starting to show. Besides ample opportunity for security breaches, ineffective or slow code, companies are also realizing that the question of ownership and maintainability is a significant one. Software has always been a liability, and generating more software in less time is not necessarily the cure for that.

However, AI does play a role at Defined Once. We use it for repetitive, easily verifiable, low stakes tasks, allowing us to focus more energy on harder problems. But we also realize that code must be "owned", and that often we write such optimized code that AI just cannot keep up (yet). Combining the fact that we focus on bringing Rust to load bearing parts of the stack, quick delivery is not the goal, but thoroughness, maintainability and clarity certainly are. Together with our critical thinking, engineering experience and years of computer security education, we do find ourselves adding value that AI can simply not provide.

Economics are shifting

As already noted, when economic uncertainty hits and costs need to be cut, contract workers are often the first ones to go. I would argue that this is why working with a collective like ours is such a benefit: scaling up or down the work is convenient and comes at a low risk. At the same time, these market conditions do pose a threat to the operation and viability of our business.

On the bright side, we sse that the demand for Rust engineers is ever-increasing. Where Rust previously was perceived as scary, unnecessarily complex and sometimes pretentious, more and more companies realize that the type system, borrow checker and helpful compiler messages do in fact reduce the number of fatal security incidents. Many now see that Rust is not only a great tool for making safer software, but that using it is also fun and attracts a completely new generation of passionate contributors. Whether you like them or not, high-impact organizations of all sizes are slowly exploring the move to Rust. The NSA has recommended Rust as one of the go-to languages for new software, and big open source projects like git are also bringing in Rust to see what it can do for them. From a business perspective, there is thus plenty to do.

This is what we also see in the software development landscape. Specifically in the Netherlands and Germany, there are only a handful of companies that are offering Rust consultancy services and training, while there is a big demand. The only reason for them to not have more clients than they do, is because they are at capacity and struggle to find capable Rust engineers to service new clients.

All in all, we are not worried about being a consultancy, dealing with AI or operating in this "new world order". We are confident in Rust becoming only more important the coming years, and our expertise and business model working out great for us and our clients. To avoid confusion, we will keep calling ourselves a consultancy but our unique offering and focus on quality will always be the goal.