.svg)
How to Reduce Bad Hires: A Hiring Improvement Guide for Growing Companies
Three months in, your top-rated marketing candidate has become a problem for the team. The resume showed impressive ad results. The interview answers were polished. But now, with a budget to manage, the new hire keeps increasing ad spend based solely on the ROAS in the advertising report. At the next hiring meeting, someone says, “We should add another interview this time.”
Before adding interviews, pinpoint where the hire fell short of what you expected. Too few applicants, disappointing performance, persistent disagreement among interviewers, and early departures call for different evidence and different fixes. Add stages without separating those problems, and you risk repeating the same mistake.
Consider a skincare brand with about 50 employees that sells through its own online store. It hires a performance marketer whose resume highlights major improvements in ad performance at a previous employer. Once on board, the marketer shifts more budget to channels showing high ROAS (ad-attributed revenue divided by ad spend). But much of that revenue comes from repeat purchases by existing customers, and almost no profit remains after discount costs. The marketer also announces a promotion before finalizing the terms with the operations lead, leaving the marketing lead to sort things out.
To understand what went wrong, revisit the evidence behind the strong hiring recommendation. Did you credit the candidate for the team’s results? Did you ask which metrics informed their decisions? Review those questions alongside the budget and decision-making authority they received after joining.
Define What a Successful Hire Means
Filling a role on time and hiring someone who delivers are different outcomes. Offer acceptance tells you whether a candidate accepted the proposal. Retention tells you whether they stayed for a defined period. Job performance shows the results and behaviors associated with their assigned work. Start your hiring review by agreeing on which outcome you want to improve so everyone addresses the same problem.
For the marketer, replace “The hire didn’t work out” with something more precise: “We wanted someone who could review ad performance and recommend the next spending plan. But we never observed whether they separated verified ad-attributed revenue from estimates or assessed profitability after discount costs.” The more specific you are about expectations and observations, the easier it becomes to decide whether to change the job posting, interview questions, work sample task, or onboarding.
Match the Evidence to the Hiring Problem
Hiring problems usually surface in one of four ways: too few applicants, a gap between interview ratings and performance on the job, persistent disagreement among interviewers, or early departures. Each points to a different set of evidence and possible causes.
| Hiring Problem | Evidence to Review First | Possible Cause | Next Step |
|---|---|---|---|
| Too few applicants | Channels that reach target candidates, actual responsibilities and terms in the posting, drop-off at each application stage | The channels, advertised role, or terms may not fit the candidates you want. | Review the posting, channels, and terms. Track reach separately from application conversion. |
| Strong interview, weak execution | Answers establishing individual contributions, realistic work sample tasks, responsibilities and authority after joining | You may lack evidence of the candidate’s own contributions or ability to execute. | Compare the experience they described with the actions they demonstrated. |
| Persistent interviewer disagreement | Differences in questions, criteria, and notes; areas not yet verified | Interviewers may use different criteria, questions, or recording practices. | Have interviewers rate independently, then compare judgments against the same evidence. |
| Early departures | Information shared before joining, assigned work, manager support, the employee’s stated reasons for leaving | Expectations about the role or working conditions may have differed, or support may have fallen short. | Review working conditions and onboarding alongside the selection process. |
Similar symptoms can have different causes. A thin applicant pool might reflect the wrong channels or an unappealing compensation package or role in the posting. If someone who excelled on a work sample task struggles after joining, check whether the task reflected the actual work and whether they received the information and support the job requires.
If interviewers keep disagreeing, start by aligning their questions and evaluation methods. The U.S. Office of Personnel Management (OPM) defines a structured interview in its structured interview guidance as asking every candidate the same predetermined questions in the same order and evaluating all responses against the same rating scale and standards. When interviewers ask different questions and score their impressions, disagreement may reveal differences in the process more than differences in the candidate.
Vague Criteria Produce Vague Evidence
“Good with numbers and takes initiative” means different things to different interviewers. Define the work the person needs to accomplish at your company, then identify the behaviors that would demonstrate their ability to do it. In its job analysis guidance, OPM describes job analysis as the foundation of all assessment and selection decisions: examining job tasks, the competencies those tasks require, and the connections between them.
Research also links job-specific assessments more strongly to later performance. In a 2022 meta-analysis published in the Journal of Applied Psychology, University of Minnesota professor Paul Sackett and colleagues reanalyzed selection research to produce revised, corrected estimates of the correlations between selection tool scores and job performance ratings. Structured interviews ranked highest at .42, followed by job knowledge tests at .40 and work samples at .33. Unstructured interviews came in at .19. The researchers noted that the top-performing methods all measured something specific to the job, while concluding that selection tools generally had substantially less predictive power than previously reported.
For the marketer, assessing “good with numbers” means observing what they do with ad performance data. Your criteria might include separating verified ad-attributed revenue from estimates, considering how budgets and discounts affect new customer counts and profit after discount costs, and checking inventory and discount terms with the operations lead before announcing or launching a promotion. In an experience-based interview, ask which metrics informed a campaign budget change and what they checked, with whom, before execution. For a work sample task, provide a channel-level performance table with ad spend, ad-attributed revenue, first-time customers, and coupon discount costs. Ask for a proposed budget allocation and a list of items to verify before execution. See the hiring scorecard example for a way to connect responsibilities, observable behaviors, and assessment methods.
Choose One Change for Your Next Hire
Changing every stage at once makes it hard to explain what made a difference. Choose the most likely cause, then document what you will change for the next hire and which evidence you will review. Treat this as a record of recurring gaps in your evaluations, rather than an experiment that estimates impact from a small sample.
| What to Document | Hypothetical Marketer Hiring Example |
|---|---|
| Gap between expectations and performance | Increased the budget without examining the composition of ad-attributed revenue or discount costs; announced a promotion before finalizing discount terms. |
| Possible cause | The interview covered ad performance figures, but not how the candidate interpreted them or what they verified before execution. |
| Change for the next hire | Add follow-up questions about the candidate’s own work and a task using an ad performance table. |
| Conditions to check alongside selection | Check whether onboarding explained budget approval limits and the process for coordinating promotions with the operations lead. |
| Evidence for the review | Compare the hiring scorecard, task responses, and interview notes with actual campaign and promotion execution and feedback from the marketing lead. |
A hiring review should go beyond judging one employee. Put the hiring scorecard, interview notes, and observations after joining side by side. If you verified the expected behaviors during selection but they did not carry over to the job, examine onboarding and working conditions. If you never assessed those behaviors, start with the questions and work sample tasks.
Find the Right Resource for Your Hiring Problem
If hiring criteria keep shifting, start with the template for aligning criteria with hiring managers. If you lack evidence of candidates’ abilities, use the work sample brief, sample answers, and scoring rubric and the assessment methods comparison. As more interviewers join the process, turn to the structured interview guide. If the final decision stalls, use the candidate evidence comparison table.
Consider Interview Pro to review interview conversations and the evidence in candidates’ answers, or Assessment to examine their decisions and performance in work scenarios. If the problem lies in hiring channels, compensation, or working conditions, start with resources that address those issues. Use the hiring process design guide alongside the quality-of-hire review checklist to compare selection assessments with performance after joining.
To distinguish a polished application or assignment from the candidate’s own judgment, use these verification questions for AI-assisted applications and assignments.
Frequently Asked Questions
Do More Interviews Reduce Bad Hires?
Before adding stages, identify the question each one answers. If several interviewers hear the same account of a candidate’s experience, consider whether a work sample task or more specific follow-up questions would tell you more. In the meta-analysis cited above, validity estimates differed substantially between structured and unstructured interviews: .42 versus .19.
Can One Metric Capture Hiring Success?
First, define the denominator and observation period. Filling a role, securing an accepted offer, retaining an employee, and meeting performance expectations are different outcomes. Track them separately and interpret them together.
Related Reading
- When the CEO and the Team Lead Are Looking for Different People
- Work Sample Task Example: From Brief to Answers, Rubric, and Evaluation Notes
- Hiring Assessment Methods Compared: Interviews, Work Samples, and SJTs
Sources
- OPM · Job Analysis — Explains job analysis as the foundation of assessment and selection decisions, examining tasks, competencies, and the connections between them.
- OPM · Structured Interviews — Defines structured interviews as asking the same questions in the same order and evaluating responses against the same rating scale and standards.
- Journal of Applied Psychology · Sackett et al., Revisiting Meta-Analytic Estimates of Validity in Personnel Selection (2022) — Reports validity estimates of .42 for structured interviews, .40 for job knowledge tests, .33 for work samples, and .19 for unstructured interviews, alongside the researchers’ conclusions.
Related Articles
