Ninety minutes against a real account with a real fault and a budget they may not raise. I score the judgement in writing; you hire.
About this service
The technical screen is ninety minutes against a real, anonymised advertising account with something genuinely wrong in it, and a budget the candidate is not permitted to raise. I score judgement, in writing, and hand you the evidence. You make the hire. I take no fee tied to a placement, which means my score does not move depending on whether anybody is hired.
What the fault is:
Different each round and never disclosed. A conversion counted twice since a tag was duplicated. A broad-match campaign quietly eating the brand budget. A smart bidding strategy still optimising toward a value that stopped being sent in March. The candidate works out loud. What I score is the order of operations: whether they check measurement before touching bids, whether they notice the budget constraint before proposing to spend past it, whether they can say they do not know and then say precisely how they would find out.
The second exercise:
A one-page brief with a contradiction inside it. Halve cost per acquisition, hold volume, no changes to the landing page, answer needed Friday. A candidate who accepts it and starts planning scores low. The role you are hiring for includes telling a director that the brief is not possible and naming which of the three constraints has to move. Someone who cannot do that in an interview will not do it in month four, when it costs them something.
The scorecard:
Five weighted dimensions written against your role, not lifted from a template, each with a description of what a low and a high score look like in your context. If the hire will run a foundation's Google Ad Grant, keyword discipline under the account-level click-through floor and the single-word keyword prohibition is one of the five. If the hire owns an edtech's enrolment peak, planning against fixed calendar dates is one of the five. The scorecard stays with you and outlives this search.
What you receive for each candidate:
Two pages. A score by dimension, a short transcript of what they said at the moment they hit the fault, and a closing paragraph naming what I would worry about if you hired this person. No strong-yes theatre with nothing attached to it, and no candidate ranked without a written reason you can argue with.
The conversation before any screening:
Often the role scope needs rewriting first. A performance marketing manager expected to also own the CRM, the website and the brand is three jobs, the market prices it as three jobs, and you will interview for a year. I will say so. I have also told clients that the honest answer was six months of an experienced contractor rather than a junior hire plus a training plan nobody has time to deliver.
What I do not do:
No sourcing, no candidate databases, no outreach, no scheduling. No reference calls, no salary negotiation, no offer management. No placement fee, at any point, for any outcome. I am not a recruiter and I will not become the cheap substitute for one.
How screens run:
Remote, in English or Turkish at the candidate's choice, recorded only with their written consent. Candidates are told in advance what the exercise involves and that they may not use their current employer's account or data. Anybody who offers to is scored down and told why, on the call.
Who this is not for:
Agencies screening at volume to fill a bench. Companies without an approved salary band, where screening produces candidates you cannot make an offer to. Anyone who wants me to make the decision. The person who will manage this hire should be the person who chose them, and should have to explain why.