An unreachable reference leaves a question, not a verdict

## Ask what the unanswered request was meant to establish

An unanswered reference request can mean many things. A company may have closed, records may have moved, the named contact may have left or the available administrator may be unable to confirm an old period of employment. None of those explanations proves the candidate's account. Equally, the absence of a reply does not establish that the account is false. A recruitment process needs to hold that distinction long enough to decide what evidence is required next. Otherwise, a failure in the route used to obtain evidence becomes a conclusion about the person being assessed.

At the fictional Lorqavi, a review found 42 candidates whose previous employment could not be checked through the usual contact. The files were all labelled incomplete, but the circumstances differed. Managers saw vacancies remaining open. The verification team saw a standard it was expected to uphold. Candidates saw repeated requests for something they could not necessarily make a third party provide. Each perspective contained a real concern. My role was to help define the unresolved question rather than allow the shared status label to conceal it. We needed to know which fact the missing response would have established and why that fact mattered to the proposed role.

Employment verification is not the same as a full assessment of suitability. A record may establish dates and a position without telling us whether someone can do the new job. A favourable conversation with a former manager may offer a useful view while still requiring corroboration of essential qualifications. If those purposes are mixed together, the organisation can place too much confidence in a familiar reference and too much suspicion on an unavailable one. I wanted colleagues to describe the evidence in its proper scope. A reliable date record supports a date conclusion; it does not automatically support every conclusion needed before hiring.

The same discipline applies to uncertainty. A file can remain unresolved without the person being described as dishonest. That may still affect what the company can responsibly decide, depending on the role and the materiality of the missing fact. The process should explain that consequence rather than disguise it as a moral judgement. It should also avoid giving candidates false reassurance that an alternative document will necessarily result in an offer. The organisation owes a relevant and proportionate process, not a promise that every difficulty will be resolved in the applicant's preferred way. Clear language helps preserve both standards and respect.

## Change the route without abandoning the standard

We worked with appropriate advisers to identify alternative records and corroboration for the particular gaps, using consent and proportionate handling. This was not an invitation to accept any document simply because recruitment was taking too long. Relevant qualifications and essential role requirements remained in force. The question was whether another credible route could establish the fact that the former-employer response was meant to establish. Different cases needed different evidence. A single replacement form would have risked recreating the same problem under a new name, especially where the underlying reason for the missing response varied.

The verification team reasonably worried that managers would use the revised process to press for convenient completion. We therefore kept decision responsibility clear and recorded what was established, what remained uncertain and the basis for accepting the evidence. A manager's desire to fill a vacancy did not decide whether a check was sufficient. Nor did HR claim expertise it did not possess in evaluating specialised qualifications or legal questions. Those matters needed appropriate professional review. A process becomes weaker when speed replaces judgement; it can become stronger when the evidence route is reconsidered carefully and the basis of the conclusion becomes more explicit.

Candidates also needed to understand why a further item was being requested. Repeatedly asking for records without explaining their purpose can make a legitimate check feel arbitrary and can encourage unnecessary disclosure. We wanted the request to be linked to the unresolved fact, with appropriate handling and no invitation to send unrelated private material. The revised route did not require people to expose their entire personal history to compensate for a missing employer response. The company needed enough evidence for the relevant decision. Collecting more material simply because it was available would have added burden without necessarily improving the conclusion.

The review cost ₹14 lakh in additional verification and process work. That expense belonged in the account, alongside the time spent by candidates and internal colleagues. We did not claim that a more thoughtful process would always be quicker or cheaper. Some cases took longer because the evidence needed careful examination. The value was a decision more closely connected to the fact being established. Where a standard route worked well, it could continue. Where it failed for reasons unrelated to the candidate's conduct, the organisation had a way to examine another route without quietly lowering the requirement or leaving the file in endless repetition.

## Keep verification and appointment as separate outcomes

Of the 42 affected cases, 28 were resolved through the accepted evidence route, nine remained unresolved and five candidates withdrew. Seven of the 28 were later appointed after the remaining selection steps. The other 21 were not selected for those vacancies. Reporting all of those figures matters. Twenty-eight completed checks are not 28 hires, and nine unresolved checks are not nine findings of dishonesty. Five withdrawals also cannot be assigned a single motive without evidence. The process outcome should retain the distinctions that management needed while deciding, rather than compress them into a success figure for the final presentation.

The seven appointments were not proof that the revised route predicted performance better than the previous process. That would require a different evaluation over time, with attention to the many other factors affecting performance and retention. The immediate evidence was narrower: we established relevant employment facts in cases where the original contact route had failed, then completed the remaining selection decisions. A board or executive committee should welcome that qualified account. It is more informative than a dramatic claim about talent unlocked by changing a procedure, especially when some cases remain unresolved and some verified candidates are not suitable for the available roles.

For directors, the useful oversight question concerns the organisation's approach to evidence. Does management know what each check is meant to establish? Can it distinguish an unavailable source from an adverse finding? Are exceptions based on an alternative evidential route or simply on pressure from a hiring manager? Are candidates told what remains unresolved without being misled about the likely outcome? These questions do not require a board to adjudicate individual applications. They help directors assess whether the employment process is capable of reasoned decisions, including decisions that remain cautious when the available evidence is insufficient.

I began my career maintaining personnel files, so I understand the appeal of a completed checklist. Completion gives a busy team a clear stopping point and supports consistent administration. The experience at Lorqavi taught me to ask one further question: complete for what purpose? A process should establish relevant facts, preserve justified uncertainty and support the next decision. When the usual source cannot respond, the organisation should examine the question that remains instead of turning silence into a verdict. That approach is demanding because it requires judgement. It is also more faithful to the reason the check existed in the first place.