Buying hiring software and building it from scratch are not the only two options. A third one is to buy a platform outright and build your own layer on top: your workflow, your rubric, your branding. AI has made that middle path cheap enough for a small team to take seriously, and it is why GetHirePlus sells a lifetime licence alongside its subscription plans.
That is the short answer. The rest of this post is why we think the ground is moving in that direction, and who the licence is actually for.
Hardware got personal. Software went the other way.
Computers started out as machines you booked time on. Then they moved onto a desk, then into a bag, then into a pocket. Every step made the machine more personal, more yours, more likely to be configured the way you like it.
Software moved in the opposite direction over the same period. It went from something you bought on a disc and owned to something you rent by the seat, that runs on someone else's computer, that changes when a product manager somewhere decides it should, and that you cannot modify even when the change you want is small and obvious.
A few people noticed this early. In February 2020, Robin Sloan wrote An app can be a home-cooked meal about a messaging app he built for his family: "I made a messaging app for, and with, my family. It is ruthlessly simple; we love it; no one else will ever use it." His framing was that software does not have to scale to be worth making. "I am the programming equivalent of a home cook."
Four months later, Matt Webb picked up the thread in Personal software vs factory-produced software, asking what separates software built for a handful of people from a throwaway script. His answer was that the good ones feel finished: "They're packaged. They don't feel temporary."
Both pieces had the same catch. In 2020, writing software for four people meant being a programmer. That was the whole constraint.
What changed is the cost of the custom layer
The constraint is softer now. Someone who can describe precisely what they want can get a working change without opening a ticket, booking a contractor or waiting for a release train. Not every change, and not without review, but enough of them that the arithmetic is different.
Research groups have been circling this for a while. In June 2025, Ink & Switch published Malleable software, which argues for "tools that users can reshape with minimal friction to suit their unique needs". Their line about where the reshaping happens is the one worth sitting with: "Adaptation happens at the point of use, not through engineering teams at distant corporations." They are careful to say this is not a world where everyone writes code. It is a world where modification stops being exceptional.
Applied to business software, that is a quiet but real change. The expensive part of custom software was never the idea. It was the months between the idea and something that runs.
Personal servers are probably the next step
If software is becoming personal again, the machine it runs on is the next question. The early signs are there.
The 2026 Self-Host User Survey, run by Ethan Sholly at selfh.st and published on 12 November 2026, drew 6,221 completed responses, which it reports as a 52% increase on the previous year. That is a community survey rather than a representative sample of anything, so treat it as a measure of interest rather than of installed base. Interest is still worth watching when it moves that fast.
On the enterprise side, Flexera's 2025 State of the Cloud Report, based on more than 750 cloud decision-makers and published in March 2025, found that 21% of workloads had been repatriated from public cloud. It is worth being precise about what that is and is not: it is selective movement of particular workloads, not an exodus, and public cloud spending kept growing through the same period. The direction is interesting; the magnitude is modest.
Put the two together and the shape is a world where more organisations run more of their own software on infrastructure they control, for reasons that are partly cost, partly data rules and partly a preference for not being a tenant.
Which way does the buy-versus-build lever bend?
It used to be a real fork. Buy, and accept someone else's defaults forever. Build, and lose a year to work that has nothing to do with why you are in business.
The middle has become passable, and we think the lever now bends towards build. Not build everything. Build the part that is yours.
The parts not worth building are the ones every hiring product needs and nobody chooses a vendor for: the applicant tracking plumbing, the interview infrastructure, telephony, scheduling, scorecards, the audit record that proves who decided what. These are solved, dull and slow to get right.
The parts worth building are the ones that make a client pick you: the rubric for the roles you actually fill, the workflow your team runs, the integrations with the systems your clients already use, and your name on the product rather than someone else's.
What that looks like with GetHirePlus
Alongside the subscription plans, GetHirePlus is available as a lifetime licence. You can run it self-hosted, on your own premises, or on our infrastructure, whichever fits your security and data rules. You can customise it to your workflows and your branding.
Two things to be clear about, because they decide whether this is a good idea for you. The licence covers the platform as it stands on the day of sale, so the roadmap you get afterwards is the one you build. And if you host it, you own the hosting: uptime, backups, upgrades, the lot.
It is priced per engagement rather than from a price list.
Who this is actually for
The clearest case is someone starting a hiring business rather than a software business.
A recruiter or a small consultancy can buy the licence, spend a few weeks on the parts specific to their niche, and go to market as a hiring service. You sell filled roles, which is what clients buy anyway. The software sits underneath as your margin and your speed, not as the thing you are trying to sell. You are not competing with ATS vendors on features, and you are not explaining your pricing page to procurement.
Be honest with yourself about what that business is. It is a services business with a software advantage. The hard parts are still finding clients and delivering well. What the licence removes is the eighteen months you would otherwise spend building the boring half, and the per-seat bill that eats your margin on every client you add.
Paying for it
For most people reading this, the licence is not priced to break the bank. We would rather have it in the hands of someone who will build something on it than hold out for a number.
Payment terms are flexible. None of this is a published price list, and it is not a form to fill in. Talk to us and we will work out something that fits what you are trying to do.
When you should not do this
Most teams should use the subscription, and we will tell you so.
If you do not have a specific thing you need to change, a licence hands you an obligation without a benefit. If nobody on your side is going to maintain the customisation, it will rot. If you want a vendor who ships you new features every quarter, a licence that covers the platform as it stands today is the wrong shape and you will be annoyed in a year.
Buy the platform when you have something particular to build on it. Otherwise start on Starter, pay for the interview minutes you use, and keep your attention on hiring.
Sources
Robin Sloan, "An app can be a home-cooked meal", February 2020
Matt Webb, "Personal software vs factory-produced software", Interconnected, 18 June 2020
Ethan Sholly, 2026 Self-Host User Survey results, selfh.st, 12 November 2026