The document we use to hire senior people was designed to rule them out. We keep using it to rule them in.

Last week I wrote about what happens after the signature: six months hiring a director, six days landing them, and a year-one failure that gets blamed on the hire when it was really the landing. A lot of you replied. Several asked the obvious next question. If the capability was usually there, what was the process actually measuring for six months?

This week gave me the answer in three phone calls. None of them were about CVs. All of them were about CVs.

The brief that contradicts itself

On Tuesday I spoke to a main contractor growing their UK business from around £200m to half a billion over the next few years, most of it on tight, technically demanding programmes for clients who inspect daily. Their UK MD is currently on site with his sleeves rolled up on a large job and needs a number two, fast.

The brief started the way most do. "We like square pegs for square holes." People from similar companies, a structured CV, a structured career, good names on it.

Then came the bit that matters, and it came with a sigh. The people with those CVs, he said, tend to be "very black and white", not as hands-on, not in the weeds. They sit in an office and point. The programme has no margin in it; the client is all over safety and cost every day. "It's not for the faint hearted."

So the filter he described in sentence one selects precisely the behaviour he rejected in sentence two. The CV tells you which companies someone has worked for and which titles they held. It tells you nothing about whether they will drive to site on a Sunday, whether they will take a difficult conversation with a project director rather than email it, or whether they have the personality to hold their own with a demanding client's delivery team. And those are the things that will decide whether this hire works.

He knew that. He said it himself. The brief just had not caught up with the knowledge.

The candidate who sold himself off his LinkedIn

The same afternoon I was on a video call with an information lead. Ten years running BIM and CAD at a consultancy, four years at a tier-one on a large hospital programme, UK citizen, wants to work on schemes with some social value in them. Exactly the profile a couple of healthcare programmes I am working with need right now.

Two things he said stuck with me.

The first: "You've probably seen my LinkedIn, which summarises most of what I do." It does. It is a clean, accurate summary of what he has done. It says nothing about why he moved from consultancy to contracting, what he learned from a job that lost the company tens of millions, or why social value matters enough to him to shape where he works next. Those were the things that told me he was right for the role.

The second: when I asked where he was pitching himself, he said £90k, because that is what he is on. For the role we were discussing, the market is £125k to £130k. He had priced the role off his payslip rather than off the problem it solves. A CV does that to people. It is a historical document, so it anchors you to your history, including your last salary.

I did not ask him for an updated CV. I asked him to talk into his phone for five minutes about the projects that mattered and why, and send me the transcript. That will tell me more than two pages of bullet points ever could.

The man with a thousand people and the same problem

The day before, I had spoken to the global digital lead at a major built environment consultancy. He runs roughly a thousand people across design tools, internal digital services and client-facing products, and came into the sector from semiconductors, where everything is measured and nothing is guessed.

I asked him a simple question: with a thousand people, how do you find the right person for a new project, and how do you spot the person doing one job who has the skills for another?

His answer could have been about CVs, because that is effectively what a skills map is. The traditional competence map, he said, relies on people maintaining it themselves. "It's biased and quite often not up to date", because people are busy doing the actual work. It becomes fragmented. And, he added carefully, people tend to define skills that "don't really exist".

Their answer is to stop asking people to describe themselves and instead derive the skills map from the work they have actually done, then ask the individual to confirm it. The system does the work; the person checks it.

Read that back as a recruitment process. The CV is a self-maintained skills map. It is biased because it is written by its subject. It is out of date because nobody updates it until they need it. It is fragmented because it is organised by employer rather than by capability. And the bits that matter most, the judgement and behaviour and values, are the bits it is least able to carry.

What the CV is actually for

I am not arguing for abolishing the CV. It is very good at one job: ruling people out. Wrong sector, wrong level, wrong geography, wrong kind of project. It does that quickly and fairly and it always will.

Where it fails is ruling people in. The six-month director search I described last week spends most of its time reading documents designed for exclusion and trying to extract from them a decision about inclusion. Then everyone is surprised when the person who was perfect on paper turns out to be the office-based pointer the client said he did not want.

Two numbers from the same week sharpen the point. A change lead on a publicly funded mega-programme told me the role her programme most needs is a pre-construction director who can talk to a trust that has never built anything, and that they keep being sent delivery people who cannot. On the same programme, a senior project manager was put into a project director seat because he was cheaper; I had told them at the outset he was not a project director, and he is not. Neither of those decisions was a CV problem. Both CVs were fine. They were both problems of reading the wrong document for the question being asked.

What I do instead, and what you can borrow

Before any name goes to a client, we agree ten non-negotiables and five nice-to-haves, in person, usually in a room for an hour or more. Not "fifteen years' experience at a tier-one". Things you can evidence: has run a programme with no float against a client who inspects daily; has moved a team onto a new site in under three months; has told an MD no and been proven right.

Then every conversation with a leader is the same question, asked many ways: what have you got that shows you can do this? Where is the example? Evidence, not adjectives. Often the best evidence is something they did ten years ago and left off the CV because they had stopped thinking it was important.

The submission that goes to the client is not organised by employer. It is organised by the three reasons this person fits the non-negotiables, what drives them, how they operate under pressure, and anything in the background that might surface later. I would rather the client hears that from me than finds it out in month four.

If you are hiring:

  • Write the spec as behaviours and evidence, not as a wish list of employers.
  • Ask every candidate the same "show me" questions.
  • Book the landing conversations before the offer, because by then you will actually know the person you are landing.

If you are the candidate:

  • Do not open with the CV. Open with why.
  • Price the role, not your payslip.
  • When a recruiter asks you to talk rather than write, talk. The transcript will say things about you that you would never have typed.

The landing post ended with a diary. This one ends with a document. Open the last job spec you issued, or the last CV you sent, and ask whether it could ever have told anyone the thing that actually decided the hire. If the answer is no, you already know what to change.

The one-page spec I use, with a worked example, is here: The ten non-negotiables template.