# Caregiver Recruiting: A Worked Example

> A walkthrough of a multi-site home-care staffing engagement: what gets asked at intake, which channels get worked in which order, and what the shortlist arrives looking like.

Author: Kariaa
Published: 2026-07-26
Category: Case Studies
Reading time: 10 min read
Canonical URL: https://www.kariaa.com/reports/caregiver-recruiting-worked-example

---

import { FlowSteps, Timeline, RankedList, CompareColumns, Takeaways, FAQ } from "@/lib/report-charts";

A home-care agency operating across three neighborhoods has open caregiver
capacity at all three. Client demand is not the constraint. Coverage is. The
office manager who would run the hiring also runs scheduling, payroll and
complaints, and the agency has been posting to the same two job boards for
several months.

This is the most common shape of brief in home care, and almost every part of the
work happens before a single candidate is contacted.

## Intake: the questions that decide the search

The intake conversation is not a form. Its purpose is to find the constraints
that will determine whether the search is possible as specified, and it usually
changes the specification.

<RankedList
  items={[
    {
      label: "What exactly is the schedule, per site?",
      note: "Not the client hours as they arrive, but the shift a person could actually build a week around. This single question changes more searches than any other.",
    },
    {
      label: "What is the real pay range, including travel?",
      note: "Advertised rate and take-home rate diverge sharply in home care once unpaid travel between clients is counted. The candidate compares the second.",
    },
    {
      label: "Which requirements are legal, and which are preference?",
      note: "Certification is usually a genuine requirement. Years of experience usually is not. Every preference written as a rule removes people who could do the job.",
    },
    {
      label: "How is a lapsed certification treated?",
      note: "Decided at intake rather than case by case at three in the afternoon, because an experienced caregiver with an expired certificate is a different candidate from someone who never held one.",
    },
    {
      label: "Who replies, and how fast?",
      note: "The agency's own response time bounds everything downstream. A shortlist delivered to a mailbox nobody reads for four days produces the same result as no shortlist.",
    },
  ]}
  caption="Intake questions, ordered by how often the answer changes the search"
/>

The first question is the one that most often reshapes the brief. Home-care hours
arrive fragmented because that is how client need arrives, and agencies advertise
them in the same shape they arrive. A candidate reading three hours in the
morning and two in the evening is being asked to accept an unpredictable week at
a rate comparable to warehouse work.

**Composing two clients into one coherent block is operational effort that costs
nothing per applicant and changes the size of the available market.** Where it is
possible, doing it before advertising is worth more than any sourcing decision
made afterwards.

## Writing the role so it can be answered

What goes out is deliberately narrow: the specific days and hours, the two named
neighborhoods, the actual hourly range, the travel and mileage arrangement, and
the certification requirement stated with whatever flexibility genuinely exists.

The instinct is to widen the post to catch more people. It reliably does the
opposite. A vague post produces applications from people who have not
established whether they can take the job, and they discover the answer later,
at a scheduled interview slot with a coordinator waiting. That mechanism is
set out in [How to Reduce Interview
No-Shows](/reports/how-to-reduce-interview-no-shows).

<CompareColumns
  left={{
    title: "What the agency had been posting",
    items: [
      { label: "Flexible hours available" },
      { label: "Competitive pay" },
      { label: "Various locations across the metro" },
      { label: "CNA certification required" },
      { label: "Experience preferred" },
    ],
  }}
  right={{
    title: "What goes out instead",
    items: [
      { label: "Tue to Sat, 7am to 3pm, one client block" },
      { label: "The actual hourly range, travel stated" },
      { label: "Two named neighborhoods" },
      { label: "Certification, and how lapse is handled" },
      { label: "Only the requirements that cause a no" },
    ],
  }}
  caption="The right column is harder to write because it requires decisions to exist"
/>

## Sourcing: the boards are the last channel, not the first

The agency had been running one channel. A search of this kind runs several
simultaneously, in rough order of how many qualified people they reach per unit
of effort.

<FlowSteps
  steps={[
    { label: "Existing staff, asked specifically", sub: "Who did you train with" },
    { label: "Past applicants", sub: "Already qualified, never recontacted" },
    { label: "Training programs", sub: "Before market entry" },
    { label: "Community channels", sub: "Language-specific" },
    { label: "Boards", sub: "One input among several" },
    { label: "Direct outreach", sub: "People not looking" },
  ]}
  note="The first two cost almost nothing and are the two most commonly left unworked."
  caption="Channel order for a caregiver search"
/>

Two of these deserve emphasis because they are free and routinely skipped.

**A specific referral ask outperforms a general one by a wide margin.** Not *do
you know anyone*, which invites a search of memory and produces nothing, but *who
did you train with*, and *who did you work alongside at your last agency*. Care
workers know care workers, and the specific question retrieves names the general
one does not.

**Past applicants are the cheapest qualified pipeline in the building.** Nearly
every agency has people who applied for an earlier opening and were reasonable
but not selected, or who declined a schedule that has since changed. Most are
never contacted again.

The channel work is examined at more length in [How to Hire Caregivers and
CNAs](/reports/how-to-hire-caregivers-and-cnas).

## Screening: per requirement, with the evidence attached

Candidates who come through Kariaa complete one verified profile of eight fields,
in a fixed order set by code rather than left to the flow of a conversation:
legal first name, legal last name, phone, location, occupation, email, resume,
work authorization. It is completed once, not per application.

Applications are then checked against the requirements set at intake. Each
requirement is marked passed or failed individually, with the evidence that
produced the verdict and a confidence level, alongside an advisory fraud read.

**The per-requirement result is the part that matters operationally.** A single
overall score would collapse exactly the distinction intake worked to establish:
that a lapsed certification and an absent one are different facts, and that one
of them is recoverable inside the hiring timeline. Seeing which requirement
failed, and how confident the check was, is what lets a coordinator decide
whether to open the file rather than trust a number.

The shortlist arrives as a table with the candidate details, status, a checks
column and a fraud column, sortable and exportable. Moving someone forward is one
of four next steps: a video interview, a set of questions, a booking, or a
message.

## The handoff, which is where searches are lost

<Timeline
  events={[
    {
      year: "Intake",
      label: "Schedule composed, requirements separated",
      note: "The largest single lever, and it happens before anything is advertised.",
      highlight: true,
    },
    {
      year: "Launch",
      label: "Specific post, several channels opened at once",
      note: "Referrals and past applicants worked first because they are free and fastest.",
    },
    {
      year: "Rolling",
      label: "Applications checked per requirement",
      note: "Evidence and confidence attached to each verdict. Candidates reply in WhatsApp or Telegram, in the language they wrote in.",
      highlight: true,
    },
    {
      year: "Shortlist",
      label: "Sortable table handed to the agency",
      note: "With the credential expiry dates recorded as dates, not as ticks.",
    },
    {
      year: "Pre-start",
      label: "The interval nobody owns",
      note: "Between acceptance and the first shift. A named owner and one scheduled contact before day one, confirming where, when and what to bring.",
      highlight: true,
    },
  ]}
  caption="Where the effort sits across a caregiver search"
/>

The final row is the one that most often undoes good work. A search that produced
acceptances and then went quiet for nine days produces first-shift no-shows, and
those get recorded as candidate unreliability rather than as an unowned interval.

## What makes searches of this shape fail

Three things, in order of frequency.

The schedule is advertised in the shape client need arrived rather than in a
shape a person can live on, so the market for the role is a fraction of the
market for the work.

Only one channel is worked, so the agency measures the most contested slice of
the market and concludes there is a shortage. That distinction is the difference
between a supply problem and a reach problem, and they have entirely different
answers.

Response is slow at the agency's end. A shortlist is a perishable asset. The
candidates on it are also on other people's lists, and the engagement cannot
compensate for four days of silence after handover.

<Takeaways
  items={[
    "Most of a caregiver search is decided at intake, before anything is advertised.",
    "Composing fragmented client hours into a coherent shift costs nothing per applicant and changes the size of the available market.",
    "Publish only requirements that would actually cause a rejection. Preferences written as rules remove people who could do the job.",
    "Referrals asked specifically and previously rejected applicants are the two cheapest channels, and the two most commonly left unworked.",
    "Per-requirement checks with evidence and confidence preserve distinctions a single score destroys, such as lapsed versus absent certification.",
    "The interval between acceptance and first shift needs a named owner, or the search loses people it already won.",
  ]}
/>

<FAQ
  items={[
    {
      question: "How long does it take to hire caregivers through a recruiting service?",
      answer: "It depends on the schedule, the pay range and the market. What can be said is where the time goes: reach is usually the constraint rather than screening, and a role open for months is almost always failing before the first qualified candidate appears rather than after.",
    },
    {
      question: "What does a recruiting engagement for caregivers actually involve?",
      answer: "An intake conversation that establishes the real schedule, the real pay including travel, and which requirements are legal rather than preference. Then a specific job post, several sourcing channels worked in parallel rather than a single board, screening against the requirements set at intake with the evidence attached, and a shortlist handed over as a sortable table.",
    },
    {
      question: "Why does composing the schedule matter more than the pay rate?",
      answer: "Because care work competes with retail and warehouse work at comparable pay, and predictability is the axis on which it loses. A candidate offered three hours in the morning and two in the evening is being asked to accept an unpredictable week for a comparable rate. Combining clients into one coherent block is operational effort that costs nothing per applicant.",
    },
    {
      question: "What does a caregiver recruiting engagement commit to?",
      answer: "The work itself: intake that establishes the real schedule and requirements, sourcing across multiple channels rather than a single board, screening against the requirements agreed at intake, and a shortlist handed over with the evidence behind each check visible. Hiring outcomes depend on the market and the role as much as on the search.",
    },
    {
      question: "How are candidates contacted during the search?",
      answer: "Through WhatsApp or Telegram rather than email, because a large share of this workforce answers messages and does not read email. The language of each incoming message is detected per message rather than set once as a profile preference, so a candidate who writes in Spanish is answered in Spanish.",
    },
  ]}
/>

---

*Written by the team at **Kariaa**. Learn more at [www.kariaa.com](https://www.kariaa.com).*