Two Job Postings for One Role: Why the Blunt Version Filled the Seat Faster

Two versions of the same job posting went out from one shop last spring, a week apart, for a residential service technician on the same truck, in the same territory, at the same pay. The first was written by the person who handles marketing and the second by the lead technician who would end up sharing a parts order with whoever answered it. The polished version pulled roughly four times as many applications, and the technician's version produced the hire. The gap between those two outcomes is most of what there is to say about how jobs actually get advertised.

The Version Written to Attract, and What It Leaves Unanswered

The polished draft opened with a growing team and a culture of doing things right, described the pay as competitive, promised variety and development and the chance to be part of something, and asked for somebody motivated, reliable and eager to learn. Nothing in it was false, and nothing in it was information either. Read it as an applicant with a current job and a mortgage, and you cannot answer the three questions you brought to the page: what am I doing at ten o'clock on a Tuesday morning, what does this pay, and which part of the work do people quit over.

A posting that answers none of those collects applications from everyone broadly available on a Thursday evening, because nobody has been handed a reason to rule themselves out. The stack looks like proof that the posting worked. Sorting it costs a week of somebody's working life, and the sorting gets done on guesswork, since the posting never established what the job needs. Volume is the metric easiest to report and the one that correlates least with whether the seat is still filled in March.

The Version Written to Talk People Out of Applying

The technician's draft was shorter, duller and considerably more specific. It said the work was residential service calls, six to nine a day, mostly diagnostic, in a territory that runs to older housing with panels three previous people have already worked on badly. It gave a pay band with a real number at each end, said where in that band a new hire starts, and named what moves somebody up it. Every line was the kind of thing a hiring manager normally holds back for the second interview, on the theory that saying it sooner would scare people off.

Then it said the two things no marketing department would sign off on. The on-call rotation is one week in four, and the phone genuinely rings around two in the morning a couple of times per rotation. Applicants whose experience is entirely new construction tend to find the first two months hard, since two of the last three people hired out of that background left inside a year. Both sentences cost applications, both were true, and the shop already knew they were true, so the only real question was whether the applicant found out in week one or week seven.

What Each Line Does to the Stack

Competitive pay and a stated band do opposite work. The word competitive removes nobody, since it stands in for a number the employer has decided not to say and experienced applicants read it exactly that way, while a band with both ends on it removes everyone whose floor sits above your ceiling before either side has spent an hour discovering that. The same pattern repeats down the page. A fast paced environment is a mood, six to nine calls a day is a workload somebody can price against their current job, must be a team player filters nobody at all, and sharing a bay and a parts order with two other technicians describes the friction that virtue was there to paper over.

The Four Lines That Do All of the Filtering

Strip a posting back to what actually works and four things are left: a pay band with both ends stated, the shape of an ordinary day in units the applicant already uses, the named difficulty written down without apology, and one sentence about what you are not willing to train. That last line saves more time on both sides than the other three together. Everything else is decoration, and decoration is expensive here, because each ornamental sentence pushes the useful ones further down a page most applicants read for about eleven seconds before deciding.

Putting the on-call rotation into numbers does a second job that has nothing to do with recruiting. Whether a callout at two in the morning is paid, and at what rate, is a wage and hour question rather than a matter of shop custom, and an employee who decides the answer is wrong takes it to the Department of Labor rather than raising it with you. A rotation written into the posting in plain figures has already opened that conversation on the record, in a document both sides kept, which is a better starting point than a disagreement eight months in.

Why the Smaller Stack Is the Cheaper Outcome

The honest version costs applications, and that is the part employers cannot stomach, since anybody inside the business measured on applicant volume will read the blunt posting as a failure for about three weeks. The counterargument is arithmetic rather than principle. A hire who leaves at week seven has cost the recruiting time twice, the onboarding once, the callbacks their unfinished work generates, and the goodwill of the two people who covered while the seat sat empty, which is a four figure number before anyone has argued about it and frequently a five figure one.

It also changes the first interview, which is where most of the saving quietly appears. With the difficulty already on the table the conversation starts on the far side of it, and nobody spends twenty minutes discovering that the applicant assumed weekends were off. That time goes instead into finding out how somebody reasons through an intermittent fault, which is the only thing an interview was ever for. The people who apply have read the hard part and applied anyway, and that self selection is worth more than any question you would have thought to ask them.

The experiment is cheap enough for any shop with a seat to fill. Post the blunt version, leave the polished one in the drawer, and if two weeks produce nobody at all you have learned something specific rather than something vague: the band is wrong for your market, or the difficulty you named is genuinely disqualifying, or the role as written is two jobs sharing one title. Eighty resumes yielding four interviews and no offer teach none of that. The shop that ran both versions now writes the way its technician did, and the stretch from posting to a signed start date is shorter than it was in the years they were drowning in applications.