Candidate

The overlap is large and the edges matter. Someone approached directly, who agrees to be considered but never fills in an application form, is a candidate and not an applicant. Depending on how a system is configured they may not appear in the funnel at all, which distorts every ratio computed from it.

Organisations that source heavily and measure only applicants will report conversion rates that describe a minority of their hiring. The fix is not a better term but a consistent definition applied to the measurement.

Funnel measures are ratios, and a ratio is only as meaningful as the definition of its denominator.

• Applications per hire looks very different if sourced candidates are included or excluded.
• Conversion at first stage depends heavily on whether unsolicited applications are counted alongside targeted outreach.
• Time to hire measured from application excludes the sourcing effort that preceded it, which can be the longest part.
• Source effectiveness comparisons break entirely if one channel enters the funnel as applicants and another as candidates.

The practical requirement is to define the entry point once, apply it to every channel, and state it alongside any number that gets reported. A conversion rate with no stated denominator invites the reader to assume the one that flatters.

It is also worth measuring from first contact rather than from application for sourced roles, since that is when the organisation started spending effort.

A sourced candidate whose details were gathered from a professional network, a referral or a conference has not volunteered their information to you in the way an applicant has.

They are nonetheless a person whose personal data the organisation now holds and processes, and the obligations that attach do not depend on whether they filled in a form. This is the population most likely to be sitting in spreadsheets, personal mailboxes and sourcing tools rather than in a system with retention rules.

• Hold sourced candidate data in the same system as applicants, so retention and access apply to both.
• Tell people what you hold when you first make contact, rather than after they express interest.
• Apply a retention period to candidates who did not progress, and to prospects who never responded at all.
• Delete on request, and be able to actually do so, which requires knowing where the data is.

The Digital Personal Data Protection Act, 2023 is the governing instrument, and its detail is dealt with in the entries covering it rather than restated here. Its substantive duties are not yet in force: notification G.S.R. 843(E) of 13 November 2025 brings sections 3 to 17, which carry the grounds for processing, notice, consent, the obligations of a data fiduciary and the data principal rights, into force on 13 May 2027. The definitions, the Data Protection Board and the penalty machinery commenced on 13 November 2025. So the practice described here is how to be ready, and good design in its own right, rather than a duty that binds today.

The most under-used source in most organisations is the people who nearly got hired.

A candidate who reached a final stage and was not selected has been assessed, has demonstrated interest, and is known to be capable of doing something close to the role. Re-approaching them for the next similar opening costs almost nothing and skips most of the funnel.

It rarely happens, for ordinary reasons: the rejection was final in tone, nobody owns the relationship afterwards, and the record is hard to search by anything except the requisition they applied to.

• Reject in a way that leaves the door open where you would genuinely re-approach, and do not where you would not.
• Record why someone was not selected in terms that will make sense to a different recruiter later.
• Make past candidates searchable by capability rather than only by the role they applied for.
• Actually contact them when a relevant role opens, rather than starting a new search from zero.

An organisation that hires the same role types repeatedly and never revisits its own near-misses is paying to rediscover people it already found.

Most people who become candidates are rejected, so the rejection experience is the organisation's most common candidate interaction and the one most likely to be described to others.

The failures are consistent and cheap to fix: no response at all, a response months later, a decision communicated without any reason, and interview processes that consume days of someone's time and end in silence.

The cost is not abstract. Rejected candidates are also customers, referrers and future applicants, and a badly handled process is described specifically and repeatedly. It also disproportionately affects the people you most wanted, since strong candidates have alternatives and compare.

Responding promptly, giving a reason where one can be given, and being honest about timelines are unglamorous and account for most of the difference between a process people speak well of and one they warn others about.