Questions
The hard ones, answered honestly.
Including the ones where the honest answer costs us the sale. In our experience it is usually the more persuasive one anyway.
About the work
We have already bought AI. Why would we need you?
That is the normal situation, and it is usually the reason people call. Most agencies have bought something and seen no change in their numbers. When that happens the cause is almost never the tool: it is that the database it runs on is full of duplicates and dead records, the systems around it do not talk to each other, and nobody agreed a number to measure before switching it on. We fix those three things. Sometimes the conclusion is that the tool you bought is fine and the problem is everything around it.
Is this not just what our ATS already does?
Often, yes, and we will tell you when it is. Most ATS vendors gate their AI features behind a higher tier, and a lot of agencies are paying for capability they have never switched on. The Tech Stack & Spend Review exists specifically to answer this question, and it regularly finds that the right move is to use what you already have, or upgrade your existing contract, rather than hire us to build something.
What if you tell us not to do anything?
Then you have spent a few thousand euros to avoid spending forty thousand, and we would call that a good outcome. Every roadmap we write has a section listing what not to bother with.
How is this different from a normal AI consultancy?
We do not sell AI. We sell recruitment operations, and we use AI where it is the right tool, which is less often than the marketing suggests. The infrastructure cost of a full automation programme for a forty-recruiter agency is roughly €10,000 to €24,000 a year. The models are the cheap part. Everything else is understanding your business, connecting your systems and getting your team to actually use it.
Will this replace our recruiters?
No, and we would be suspicious of anyone who says it will. What we automate is the admin around the job: logging, chasing, scheduling, formatting, reporting. The judgement stays with your people, partly because that is what your clients pay for and partly because the law increasingly requires it. We will not build anything that rejects a candidate without a person deciding.
Our recruiters will not use it.
That is the most likely way this fails, and it is why adoption is designed in rather than hoped for. We put at least two recruiters, including a sceptic, into the design sessions, we measure usage weekly from day one, and we kill features nobody uses. We will also ask what your team is measured and paid on, because if the incentives point the other way, no amount of good software will win.
About price and risk
Why are your prices published when nobody else's are?
Because in this market, opacity is the norm and it is not working for buyers. We would rather you knew whether you could afford us before spending an hour on a call.
Why fixed price rather than a day rate?
A day rate pays us more the longer we take, which is a strange incentive to give someone you have hired to make things faster. Fixed price means the risk of it taking longer than expected is ours.
What if the scope changes?
It becomes a new line at the price on the site. We do not do variations or hourly top-ups: that is how a fixed price quietly becomes an open-ended one.
What is the smallest thing we can start with?
A Second Opinion on something you are already considering, or a Diagnostic Lite. The first needs no systems access at all: just send us the proposal.
We are a small agency. Is this worth it?
Under about ten fee-earners, honestly, most of the build packages will not pay back: your total admin cost base is not big enough to recover a €36,000 investment. Diagnostics and advisory still make sense, and we will tell you what to buy off the shelf. We would rather say that than sell you something that does not work.
How do we know it worked?
We agree one number before we start (time to first submission, redeployment rate, approval delay, whatever matters to you), measure the baseline, and report against it. On builds over €18,000, the last 20% of the fee is not due until after the thirty-day review. And if the number does not move, we will tell you that too. Around 95% of enterprise AI projects show no measurable financial impact, mostly because nobody checked.
About data, risk and the law
Where does our candidate data go?
Ideally nowhere new. We build on a self-hosted platform inside your own cloud account, so candidate data stays inside your boundary. Where a third-party service is genuinely needed (transcription, for example) we will tell you which one, where it processes data, and what its retention terms are, before you commit.
Is this legal? What about GDPR and the AI Act?
Automating admin is straightforward. What needs care is anything that affects a candidate's chances, and the rule we build to is simple: a machine can gather, structure, rank, draft and explain; a person has to decide. That keeps you the right side of the automated-decision rules under GDPR, and it is the direction the EU AI Act is heading for employment systems from December 2027. We are not lawyers and we do not pretend to be, but we will hand your legal adviser a document they can work from rather than starting cold.
Can you make the AI reject the bad applications automatically?
No. Technically it is easy; legally it is the single clearest way to create exposure, and a human rubber-stamping the output does not fix it: regulators have said so explicitly. What we build instead ranks candidates and shows the evidence, so your recruiter sees the best twenty with the reasoning attached and makes the call. You get most of the time saving, a better shortlist, and a record of how the decision was made if anyone ever asks.
Can you automate our LinkedIn outreach?
No. LinkedIn's terms explicitly prohibit automated access and scraping, including browser plugins, and they enforce it, and it is your account that gets restricted, not ours. We will show you the compliant alternatives, and we will usually suggest making your own database searchable again, since you are already paying five figures a seat for LinkedIn partly because you cannot find people in the system you own.
Do you do payroll?
No, and we will not. Specialist platforms do it properly, and getting it wrong is a legal and financial problem nobody should be casual about. We automate everything around it: chasing timesheets, extracting the data, chasing client approvals, and leave the calculation to the people whose job it is.
About us
How big are you?
Small, deliberately. You get the person who does the work, not an account manager and a delivery team you never meet.
Do you take commission from vendors?
No. Never have, never will. It is the reason our recommendation is worth something.
What if we want to stop?
Diagnostics are refunded in full if you cancel before we start. On builds, you pay for completed milestones. Retainers have a three-month break clause. And everything we build is documented and owned by you, so it keeps working whether or not we are involved.
Who else have you done this for?
We are new, and we would rather tell you that than imply a client list we do not have. What we do have is a researched view of how recruitment agencies actually operate, which is why our first conversation will be about your desk economics rather than about AI. Our early engagements are priced accordingly, and we are happy to be judged on the diagnostic before you commit to anything larger.
Why is your catalogue so short?
It is not, but what we publish is. We have around forty packages researched and costed, and we are selling nine of them plus three bundles. A forty-item price list from a new firm is a wish list, not a service catalogue. We switch the rest on as each one has real deliveries behind it, and we will tell you plainly when something you want is in that group.