Can't Find the Right Hire? Which Requirement to Relax First

When you can't find the person you want, relax the requirement that is quickest to learn on the job and cheapest to get wrong. Most companies do the opposite. They hold on to the requirements that are easy to write down, like a language certificate or years in the industry, and quietly drop the vague ones, like judgment, that can't be taught at all.

What order should requirements be relaxed in?

Rank each requirement on how long it takes to build after joining, whether it can be built at all, and what a mismatch costs. Relax the fast and cheap ones first. Never relax the ones that can't be built and are expensive to get wrong.

That sounds obvious. In practice it gets reversed, because of how job posts are written. Japanese level and industry years are easy to quantify, so they become the hardest lines in the post. Judgment and values are hard to put into words, so they end up as a soft phrase at the bottom. When the search stalls and something has to give, the soft phrases go first. Nobody notices they're gone. The hard lines stay, because nobody wants to be the one who lowered the bar.

So the company protects what is easiest to fix and gives up what can't be fixed.

The four layers make the pattern clear. Split a person into core, patterns, equipment and context. Equipment is the one layer you can clearly build after someone joins. Core barely changes, and patterns are hard to change. Industry experience, tools and certificates all sit in equipment. Judgment is a pattern. Values sit in the core. The right relaxation order runs from the outside in.

Industry experience predicts less than you think

Industry experience is safe to relax early because it predicts surprisingly little, and what it does predict fades fast.

In a 2019 meta-analysis in Personnel Psychology, Chad Van Iddekinge and colleagues pooled 81 independent samples to test whether pre-hire work experience predicts results in a new job. The corrected correlation with job performance was 0.06 (44 samples, 11,785 people). With training performance it was 0.11. With turnover it was 0.00. Even experience directly relevant to the current job was only weakly related to outcomes. The authors concluded that experience-based screening measures generally predict future performance and turnover poorly.

There was one exception. Experience predicted performance somewhat better when workers were just starting the new job.

That exception tells you what industry experience is. It's a head start. Experienced hires ramp faster and make fewer mistakes in the first quarter. After that, the others catch up. So "five years in our industry" can go, as long as you pay for it in the first months: a ramp plan and one named person to teach. That trade usually works out.

Tools are even easier. Whether someone knows a particular CRM or project tool is a matter of a few weeks. Making it mandatory shuts out good people to save a few weeks of training.

What does "JLPT N1 required" actually screen for?

Reading and listening. It says nothing direct about speaking with a client.

The official FAQ of the Japanese-Language Proficiency Test says plainly that there is no conversation test and no composition test, and no plan to add either for now. N1 proves someone can read and follow complex Japanese. Whether they can sit across from a client and pin down what the client actually needs is outside what the certificate measures.

I learned this the hard way. I was searching for a bilingual English and Japanese IT project manager, and I insisted on native-level Japanese. Six months went by. Once I dropped that one requirement, I hired someone within a month. Looking back, "native-level Japanese" was standing in for working through requirements with Japanese clients. Business-level Japanese covers that, someone else can polish the documents, and working Japanese keeps improving on the job. "Native-level" described an imaginary person.

Once you split it, the language requirement has two tiers. The certificate and the written polish can be relaxed, built up, or covered by a partner. The tier where someone negotiates, nails down requirements and handles complaints in Japanese is the core of the role. It takes years to build, and a mismatch lands on your clients. Keep that tier.

A related waste is hiring trilingual people and then using them only to translate. I wrote about that in Your bilingual staff are not translators.

Relax what is quick to build and cheap to get wrong; keep what can't be built and is costly to miss

Hiring requirements placed by time to build after joining and cost of a mismatch. The grouping is the author's judgment framework, based on the four-layer model and Van Iddekinge et al. (2019)

Across: time to build after joining, short to long. Down: cost of a mismatch, low to high

Relax first

Tools, systems, industry jargon. Learned in weeks to months; a miss costs a little extra training.

Make it a plus

Degrees, the JLPT certificate itself, a second foreign language. Slow to build, but the role works without them.

Relax, with a ramp plan

Years in the industry, product knowledge. Experience correlates 0.06 with performance and matters mostly early. Name a teacher.

Keep

Judgment, values, and the tier of Japanese needed to work out requirements with clients. Hard to build after joining; a miss is priced in salaries.

Relax from the equipment layer inward. Protect core and patterns.

Why the unbuildable requirements stay

Because a mismatch on those is paid for after the hire, month by month, with no way to make it up.

Judgment is a pattern in the four-layer sense: how someone decides under pressure, which way they lean when information is missing. It's hard to teach after joining. Values sit in the core and barely move. If a person's line doesn't match what your company actually rewards, no training will move it.

Put numbers on it. One hire in Japan costs 1.2 to 1.5 times salary a year, counting statutory costs and, in the first year, the recruiting fee. For a role paying 8 million yen, that first year runs around 10 million yen, and a wrong hire wastes most of it, before counting the months to hire again or the bad decisions made in the meantime. Relaxing industry experience costs you a few months of coaching. Relaxing judgment costs you the 10 million.

Put the two side by side and the choice isn't hard.

My view is that the vaguer judgment and values are in the job post, the more concretely you need to probe them in the interview. Left vague, they're the first things to disappear without anyone deciding to drop them.

Decide the order before the post goes out

Decide the relaxation order before the search stalls. Three months in with no hire, the hiring manager and HR are both under pressure, and whatever is easiest to cut gets cut. That's rarely the right one.

The method I use is the talent-map part of The two maps. For each requirement, write down four things: what it actually screens for, where it can bend, how long it takes to build after joining, and how many more years it stays useful (and why it would stop being useful). Once that's done, the order mostly writes itself. The shorter the build time, the earlier it goes. Anything marked "can't be built" is your floor.

Add one more line: what you are not requiring this time. If it isn't written down, a dropped requirement creeps back within weeks. An interviewer glances at a resume and mutters, "no industry background, though."

Each requirement you relax moves the search one ring outward, from the same industry to adjacent ones. Recruiters usually haven't called the people out there yet. Whether you can use them depends on whether you can write the ramp plan. If you can't, relaxing the requirement just moves the problem to after the start date.

---