Home-Care Software Buyer’s Guide

Summary: Evaluate a platform through real workflows, state and payer fit, evidence, controls, implementation, and total cost. A feature checklist alone will not show how the system handles exceptions.

Choosing home-care software is difficult because broad product labels sound similar: scheduling, EVV, caregiver app, billing, payroll, CRM, family portal, reporting, and now AI.

The difference appears when something changes, fails, conflicts, or needs approval. Build the evaluation around those moments.

1. Define the operating scope

Document:

• care setting and service types;
• states, locations, and branches;
• private-pay, Medicaid, managed care, LTC insurance, VA, or other payer mix;
• active clients, caregivers, visits, and growth expectations;
• current scheduling, EVV, communication, billing, payroll, accounting, recruiting, and reporting systems;
• mobile, language, accessibility, and connectivity needs;
• required integrations and contract dates.

Do not accept “multi-payer” or “all 50 states” as a substitute for your exact list.

2. Choose five real workflows

Use scenarios such as:

1. A caregiver calls off four hours before a qualified visit.
2. A mobile clock works offline but the external EVV transmission fails.
3. An authorized family member needs a clear update after an exception.
4. A completed visit is missing a billing prerequisite.
5. An applicant responds after hours and needs an interview and accommodation.

Ask the vendor to run each flow from beginning to end, including roles, messages, failures, corrections, approvals, and audit history.

3. Evaluate scheduling as a decision system

Ask:

• Which criteria determine eligibility and which determine fit?
• Can we configure agency and location policy?
• Does the scheduler see why someone was recommended or excluded?
• How are open-visit messages sent, tracked, expired, and assigned?
• How does a schedule change reach the caregiver, EVV, client, billing, and payroll?

4. Verify EVV precisely

CMS states that federal law requires EVV for specified Medicaid-funded personal care and home health services requiring an in-home visit, while implementation occurs through state systems. (CMS EVV, retrieved 2026-08-11.)

Ask for a dated matrix covering your state, program, service, aggregator, verification methods, offline behavior, correction policy, transmission status, and support ownership.

5. Trace service into revenue

Take one visit and follow:

• planned service and authorization;
• delivered and verified service;
• documentation and review;
• rate and payer rules;
• invoice or claim preparation;
• submission and response;
• remittance and payment;
• caregiver pay;
• adjustments and reconciliation.

If teams must export, re-enter, or reconcile manually, identify the volume, owner, evidence, and cost.

6. Review communication and consent

For SMS, email, voice, portal, and family updates, ask:

• Who can contact whom, for what purpose, and through which consent?
• Which content is safe for the channel?
• Are delivery, response, failure, and opt-out visible?
• How is recipient authorization maintained?
• Can staff separate internal notes from external updates?
• Does automation escalate sensitive or urgent content?

7. Test the caregiver experience

Have actual caregivers or field leaders evaluate:

• schedule clarity;
• login and device requirements;
• visit start and end;
• task and note completion;
• offline behavior;
• language and accessibility;
• error recovery;
• notification load;
• support path.

Caregiver experience affects data quality, adoption, and retention. Do not evaluate it only from an administrator’s demo account.

8. Ask AI questions that expose the workflow

The 2026 ACL/NCOA review finds promising AI applications in home care but limited independent evaluation and material risks involving privacy, consent, surveillance, bias, data quality, access, and human connection. (ACL/NCOA report, retrieved 2026-08-11.)

For each AI feature, ask:

• What source does it use and preserve?
• Which user’s permissions apply?
• What decision or action remains human?
• How does it behave with missing, conflicting, or low-confidence information?
• Is generated content labeled?
• Which model providers process data, under what retention and training terms?
• How are output quality, bias, incidents, and changes monitored?

9. Review security as a scoped claim

HHS describes administrative, physical, and technical safeguards for ePHI and explains business associate responsibilities for cloud providers handling it. (HHS Security Rule, HHS cloud guidance; retrieved 2026-08-11.)

Ask for service boundaries, access controls, organization isolation, encryption, audit records, secure development, incident response, backup and recovery, retention and deletion, subprocessors, contracts, and customer responsibilities. A logo or BAA is not a complete review.

10. Price the complete change

Include:

• subscription unit and volume assumptions;
• modules and usage;
• implementation and configuration;
• data migration and validation;
• integrations and third-party charges;
• communication or AI usage;
• training and change management;
• premium support;
• parallel-run and contract overlap;
• internal time for cleanup, testing, and adoption;
• renewal, overage, and exit terms.

11. Demand proof

Objective claims should have a reasonable basis before they are advertised; the FTC’s substantiation policy is a useful baseline for asking vendors how outcome claims were produced. (FTC, retrieved 2026-08-11.)

For each metric, ask for population, baseline, time period, inclusion rules, sample size, method, and whether the result is typical.

12. Score fit, risk, and evidence separately

Use three scores instead of one:

• Fit: Does the workflow meet the agency’s need?
• Risk: What regulatory, operational, technical, or adoption exposure remains?
• Evidence: Was the capability demonstrated and documented, or only described?

A feature can fit on paper and still carry high risk or weak evidence.

The right system is not the one with the longest feature list. It is the one that can run your real work, make exceptions understandable, fit your obligations, and prove what it claims.

Next: Explore the Ethiya platform draft or use these scenarios in a demo.