Deciding what you are actually hiring for

Most bad hires were decided before anyone was interviewed, in the ten minutes somebody spent copying an old job description.

Write down what this person will do in their first six months. Specific tasks and outcomes, not a list of responsibilities. This is the document everything else comes from, and writing it honestly often reveals that you need something different from what you assumed, or that you need two part time things rather than one person.

Separate must have from nice to have, and be ruthless. Every requirement narrows the pool, and long lists of requirements filter out good candidates while barely filtering out weak ones, because weak candidates apply anyway.

Be careful with degree and year requirements. "Five years of experience" is a proxy for capability and a poor one; people vary enormously within the same number of years. In Pakistan in particular, a great deal of real capability sits with people who are self taught or who came through a bootcamp, and a degree filter removes them silently.

Decide the salary range before you advertise, and publish it. Not publishing wastes everybody's time, including several rounds of yours, and it is a large part of why candidates disengage late in a process.

Agree who decides. One person accountable, others advising. A hiring process where four people have a veto ends in the least objectionable candidate rather than the best one.

Know what you cannot offer, and be honest about it up front. Discovering it at the offer stage costs you the candidate and your reputation with them.