Why weak hiring habits persist

If a familiar hiring process keeps disappointing you, changing it should seem straightforward. Define the work more carefully, gather better evidence and stop rewarding a polished account over a relevant one. Yet those changes happen inside a business that still needs someone to start, a recruiter who has other searches to run and a candidate who cannot wait indefinitely.

The old process has an advantage before anyone argues for it: people know how to use it. The manager can request a title, the recruiter can find similar titles, and applicants know which document to send. Everyone can get moving while the harder questions remain unanswered.

That is one reason a weak process can survive. It coordinates activity even when it does not produce enough understanding. To change it, a founder needs to examine what the process asks people to do and what it rewards them for doing.

Familiar can feel like proven

In 1988, William Samuelson and Richard Zeckhauser studied status-quo bias, the tendency to favour an existing option over alternatives.¹ They were studying a broader behaviour than resume use, but their work gives us a question to ask about hiring: are we retaining this step because it serves the decision, or because removing it feels riskier than keeping it?

There may be good reasons to retain a familiar step. A qualification check can be necessary. A consistent application form can make relevant information easier to compare. Switching systems can impose real costs. The point is to examine those reasons rather than treating the age or popularity of a practice as evidence of its quality.

Organisations also learn from one another. Paul DiMaggio and Walter Powell described pressures that make organisations in a field become more alike, including imitation under uncertainty.² Their broader theory offers one possible explanation for why copying a respectable competitor can feel safer than designing a process around your own work. It leaves you to establish why your company uses a particular step.

If a hire fails after the usual interviews, it can be described as an unfortunate result from a normal process. If a new approach fails, the approach itself is more likely to be questioned. That difference can discourage experiments even when nobody is particularly satisfied with the default.

I have heard common practice described as best practice often enough to distrust the phrase. Doing something for years gives you experience of doing it. It does not, by itself, show that it helps you decide well.

What gets rewarded gets done

Hiring has a supply side and a demand side. The supply side concerns people who might do the job: finding them, reaching them and understanding what they can offer. The demand side concerns the work: the result the business needs, the conditions around it and the contribution the team is missing.

The supply side is visible. A recruiter can report how many people were contacted, how many applications arrived and how many interviews are booked. Work spent resolving a disagreement about the role may produce no new candidate to show the client that week, even when it prevents a much more expensive mistake.

This creates pressure to start with the search. A manager asks for candidates and the recruiter provides them. The manager sees people moving through stages and feels that the vacancy is being handled. Everyone can be busy before anyone has established what a good decision would require.

Fees can reinforce that pressure. When payment depends on filling a role, there is a reason to concentrate effort on reaching an accepted offer. That does not mean every recruiter on that arrangement disregards quality. Reputation, repeat business, replacement terms and professional standards matter too. The useful question is whether your particular arrangement gives someone time and responsibility to challenge the brief.

The same problem can exist in an internal team. A founder who asks only when the shortlist will arrive should not be surprised when the recruiter prioritises producing one. An employee asked to fill roles quickly may find it difficult to delay a search in order to question whether the job has been designed sensibly.

If that is your position, bring a recent example and the decision it exposed. The role may be described as setting priorities independently, while the manager expects to approve each change. Explain what each arrangement would require of a candidate and recommend one. Chapter 6 follows what happens when the person with authority disagrees.

Demand first, supply second is the correction I favour. Understand enough about the work to make the search purposeful, then keep learning as you meet people. That leaves room to revise the role as you learn, without demanding weeks of paperwork before the first contact.

If you use a recruiting partner, ask what they need to learn before searching and who will resolve unanswered questions. Ask how they will assess a candidate against the role, what they will show you when evidence conflicts and how you will review the result after the person starts. Those answers tell you more about the service than a promise of a large shortlist.

When volume makes the information worse

Applications and screening can create a self-reinforcing problem. Employers receive more material than they can examine carefully, so they introduce narrower filters. Applicants try to satisfy those filters and send more applications to improve their chances of being seen. Employers then have still more material to manage.

AI can reduce the effort needed on both sides. It can help a candidate tailor a genuine account, and it can help an employer organise relevant evidence. It can also make it easier to produce large amounts of material that look suitable while telling either side very little. The effect depends on how the tools are used.

I call this a Flywheel of Pain because a response that helps one person can make the next person's problem worse. The founder wants relief from an empty role, the recruiter needs a process they can manage, and the candidate wants a fair chance and a timely answer. Requiring a long form can save the employer interview time but consume hours across unsuccessful applicants. Sending hundreds of applications can improve a candidate's chances of contact while adding to the volume each employer must sort. These are choices people can change, even when each seems reasonable in isolation.

Neither the volume of applications nor the use of writing assistance establishes dishonesty. Check whether the account is accurate and the evidence relevant before excluding someone for the tool they used.

The candidate is dealing with your process too

I've watched friends put care into applications and get silence in return. When my brother was looking for work, we sat down together to think through how he should approach the process.

An applicant left without an answer has little basis for deciding what to change. A keyword filter may have screened them out, another candidate may have offered stronger evidence, or something inside the business may have changed. They may decide to apply more broadly or make the page resemble what they think employers want. They may stop applying to jobs they could do. Silence leaves room for the conclusion that volume matters more than care.

As an employer, you control part of this experience. You can state what is genuinely required, explain the stages, avoid asking the same information repeatedly and tell candidates when the search changes. Distinguish what you need to assess from what would merely be convenient to have, so the request for someone's time is easier to understand.

The employer has constraints too. A small business may not be able to give every applicant a detailed explanation. It can still provide a clear decision, avoid misleading claims about what will happen next and give shortlisted candidates enough information to decide whether continuing is worthwhile.

The part you can change

You do not have to redesign the labour market to improve one search. Start with the point where your process turns an assumption into an instruction.

Here is a hypothetical example. A founder tells a recruiter to find a senior account manager with several years in the same industry. Before searching, the recruiter asks what the person must accomplish. The founder describes missed handovers and clients receiving updates too late. The team describes a different problem: no one can confirm priorities until the founder approves them.

If the recruiter starts with the title, the shortlist may be full of experienced account managers who will encounter the same bottleneck. If the recruiter tests the disagreement, the business can decide whether it needs an account manager, coordination capacity or a change in authority. Chapter 9 develops a version of that decision in a complete worked example.

This pause has a cost. Someone must examine recent work, listen to the people involved and make a decision. The comparison is not between careful hiring and a process that costs nothing. It is between spending some effort before the search and discovering the same problem after making an offer.

Give that work an owner. The recruiter may lead the enquiry, but a founder or manager must have authority to settle the role's purpose and limits. Otherwise every unanswered question returns as another search criterion, and the ideal candidate becomes a person expected to compensate for decisions the business has avoided.

It also helps to agree what progress means. A shorter time to fill can be useful, especially when a vacancy imposes real costs. It should sit beside evidence about the quality of the decision, the candidate effort required and what happens after the person starts. A slow process can be poor too. Care is not measured in elapsed weeks.

Look at the rules before buying another tool

If a system screens for a degree the job does not require, the first repair is to examine the requirement. If a manager dismisses people for lacking an exact title, the first repair is to define acceptable evidence of the work. Buying more sophisticated screening may make the existing rule easier to apply without making it a better rule.

Tools can help once the purpose is clear. Consistent records can expose missing evidence, and a shared view of the role can prevent interviewers from judging different jobs. Automation can reduce repeated administration. Used this way, the software supports the enquiry.

The same is true of standardisation. Some consistency protects candidates. Asking comparable questions, giving equal access to task information and using agreed criteria can make a process fairer and easier to examine. The objection is to standardising the wrong thing, such as one career history, while leaving the actual judgement entirely personal.

Change the question you review

When a hire disappoints, ask both what the process encouraged and what you chose. A rushed brief may have reflected pressure from the founder. A narrow shortlist may have followed requirements nobody tested. An attractive interview may have received more weight than the work evidence. Naming those decisions gives you something to change.

Do not use the system as a way to remove your own responsibility, and do not make a single recruiter carry a problem the business helped create. Work through what each person knew, what they were asked to deliver and where a different decision was possible.

The next chapter follows the consequences when those decisions go wrong, including costs that never arrive as an invoice.