What a Good Hiring Process Looks Like End to End

You've just made an offer to someone who talked a great game and turned out to be wrong for the job. Now you're replaying the whole process trying to find the one step that failed. Usually it isn't one step. It's that there was no process at all — just a series of gut calls that happened to be in the same order every time.

A good hiring process isn't a rigid HR ritual. It's a short sequence of checkpoints, each one designed to catch a different kind of mistake. Skip a checkpoint and you're not saving time, you're just moving the risk further down the line, to the point where it costs a lot more to fix.

Here's what the sequence looks like, and what each step is actually for.

Step 1: Write down what the job needs before you write the ad

Most job ads get written from memory of the last person who did the role, or from a template someone found online. Before you write a word of the ad, answer three questions on a blank page:

• What will this person be doing in their first month that, if done badly, actually hurts the business?
• What's the one skill that, if missing, no amount of enthusiasm makes up for?
• What can be taught on the job, and what has to already be there?

This takes fifteen minutes and it's the step almost everyone skips. Everything downstream — the ad, the screening questions, the test, the interview — should trace back to this list. If a question in your interview doesn't map to something on this list, cut it.

Step 2: Screen for evidence, not vibes

When applications come in, you're tempted to skim for a good school, a recognisable employer name, or a CV that's laid out nicely. None of those predict whether someone can do the job. What predicts it is evidence: did they do something, with a result, that resembles what you need them to do here.

A fast filter that works: read each CV once, looking only for sentences that describe an action and an outcome ("reduced returns processing time by reorganising the queue") rather than a duty ("responsible for returns processing"). Shortlist on evidence density, not polish. It's a rougher filter than it sounds, but it's a lot better than a gut reaction to a logo.

Step 3: Test the actual skill before you interview

This is the step most small companies skip entirely, and it's the one that would have caught your last bad hire. An interview tells you how someone talks about the work. A test tells you whether they can do it.

It doesn't need to be elaborate. A half-day take-home task, a short work sample, or a structured test of the specific skill the job needs — all of these beat a conversation, because a conversation rewards confidence and a task rewards competence. If you're building this yourself, keep it short enough that a good candidate can finish it in the time you promised, and score it against a rubric you wrote before you saw a single answer.

This is also the point where a tool like AssessFit is a genuine shortcut rather than a nice-to-have — it's built so a CV can never carry more than 30% of the final score, which forces the decision back onto what someone can actually do. But whether you use a product or a spreadsheet, the principle is the same: test before you fall for someone.

Step 4: Run a structured interview

By the time you interview, you should already have signal from the CV and the test. The interview's job is to check for judgement, not to discover the skill for the first time. Ask the same core questions to every candidate for the role, in the same order, and take notes during the conversation rather than after — memory fills gaps with impressions, and impressions are exactly what you're trying to reduce.

A question worth asking regardless of role: "Tell me about a time you were wrong about something at work and what changed your mind." It's hard to fake well and it tells you more about how someone will behave under pressure than almost anything else you'll ask.

Step 5: Check references properly

Most reference checks are a waste of a phone call because they ask "would you recommend them?" — a question every referee answers yes to, because the candidate chose them. Ask instead for something specific and checkable: "What would you have wanted them to do differently in their last three months?" or "What kind of task did they need the most support with?" You're not looking for red flags, you're looking for texture — the difference between how a candidate describes themselves and how someone who worked next to them describes them.

Step 6: Score everyone the same way, then decide

Write your scoring criteria down before the first interview happens, not after you've met everyone and formed favourites. A simple 1–5 scale against three or four criteria — the skill test, the structured interview, the reference check, culture fit defined narrowly (how they handle disagreement, not whether you'd get a beer with them) — keeps the decision anchored to evidence instead of whoever interviewed best last.

The honest limit here: no scoring system removes judgement, and it shouldn't. It just makes the judgement visible and comparable, so you can explain the decision to yourself and to the person you didn't hire.

Step 7: Offer well, reject well

The offer should move fast once you've decided — good candidates have other conversations running in parallel, and a slow offer reads as low interest even when it's just admin. The rejection matters almost as much: a short, specific, human note (not a template that could apply to anyone) is one of the few things a candidate remembers about your company for years, in either direction.

Step 8: The first 90 days are still hiring

A process that ends at the signed offer isn't finished — it just moved the risk to probation. Set a real check-in at 30, 60, and 90 days with specific, written expectations for each milestone, not a vague "how's it going." If someone isn't working out, you'll know by 60 days if you were watching for the right things. If you weren't watching, you'll find out at month nine, which is a much more expensive place to find out.

A realistic timeline

For a single hire at a 10–200 person company with no dedicated recruiter, expect roughly: a week to define the role and post it, one to two weeks to collect and screen applications, a few days to run tests and shortlist, a week for interviews, a few days for references, and a fast offer. Three to five weeks total, most of it waiting on other people's calendars rather than your own decisions. The steps above don't make hiring faster. They make the time you're already spending point at the right things.