How to evaluate a housing and homelessness charity

A starting point, not a checklist. Homelessness is harder to evaluate than water or food banks, and three good questions asked well will serve you better than twenty.

What a homelessness organization actually controls

1. How people are treated. Safety, dignity, being met as a person. This decides whether someone comes back inside on a cold night and whether housing holds. You will not measure it from a report. You will notice it from being present on site, and one way in is to notice how long the frontline staff have been there. You can ask it without sounding like an auditor: who is the first person someone meets here, and how long have they been doing it. In this work, trust is built with a face, and organizations that cannot keep their people cannot keep that trust. Much of that is pay, frontline pay in particular. A funder can have a significant positive impact on this decision.
2. Whether it meets the need in front of it. Someone arrived needing something. Did they get it. And does anyone write down what was needed, what was met and what was not, so the gap is known rather than guessed and next year looks different. The outreach questions below are the practical version of this.
3. Whether it finds out what happened, and for whom. Not controlling the outcome. Caring enough to know. Checking in at one, three and five years does not prove anything was caused. It shows the organization wants the answer. Knowing whether results differ for Indigenous clients, youth or women leaving violence shows it knows its own population.
4. Whether it tells the truth and changes because of it. The ground moves here faster than in most causes. The drug supply, the policy environment and the housing market are not what they were three years ago. An organization doing what it did in 2021 is not paying attention.

The chain, and which part they own

Someone connects with a worker they come to trust. They get somewhere safe to sleep. They get support with whatever is making things hard. They move into a home. They keep it. They build a life inside it. Almost no organization is with a person for all of that. Ask which part of this do you own, and what do you measure about it. Then find the segments that match. Most organizations touch two or three.

Outreach

• Were you able to meet the need the person had when you met them, and where you could not, do you know why?
• Of the referrals you make, what share does the person actually reach? A walk to the clinic and a phone number on a card are both called referrals.

Good: knows how often the need was met, what the gaps were, and how often a referral landed. Weak: counts contacts.

Emergency shelter

• Are people safe here, and what happens after an incident?
• Would a guest send someone else here?
• Does everyone leave with a next step, even when it is not housing?

Good: talks about safety and care before beds. Weak: reports occupancy, which is always high, or turnaways, which measure demand rather than the shelter.

Stabilization

Addiction, mental health and health support inside a homelessness program.

• Is support offered without conditions, or is it a condition of housing?
• Do you know how people are doing a year or more on, and what do you do with that?

Good: tracks people over time and treats housing stability as the outcome, not sobriety. For the standalone question, see how to evaluate a mental health charity.

Housing

• Of the people you housed last year, what share are still housed, what share did you lose track of, and how does that compare with programs like yours?
• What share have income support in place, and their ID?

Good: states the follow-up window and knows the income-support and ID numbers, because that is what a caseworker's week is made of. Weak: near-perfect retention on a small group with no window stated.

Prevention

Working with people who are housed and close to losing it.

• Of the households you helped, what share were still housed a year later?
• Compared to when you started with them, can people cover the rent and put food on the table?

Good: has a starting point and checks against it. Weak: counts households helped.

Where impact actually lives

No single organization can claim it changed a life. Too many other forces act on one person. So the honest question about impact is asked of the community. A By-Name list is real-time data on every known person experiencing homelessness in a community, shared across agencies so people are prioritized by need. If chronic homelessness in a city is falling, something is working, no one charity can claim it, and all of them contributed. Ask whether the organization is on its community's list. A yes with the community named is a good answer. A reasoned no is an answer.

The thing no question on this page reaches is what it feels like to be on the other side of the desk. Met as a person with a history, or filed as a case. Where trust has usually been broken before, that may decide whether the housing holds. Ask the organization to show you. Better, ask someone who has been through its doors.

Once the program holds up, check the organization behind it.

Questions people also ask

What is the single best question to ask a charity that helps homeless people?

Which part of the chain you own, and what you measure about it.

Does Housing First work?

On housing stability, yes. In the Canadian trial, high-needs participants spent 85% of days stably housed over six years against 60% for standard services. On substance use and quality of life, the same trial found no measurable difference at long-term follow-up.

Is a high turnaway number a bad sign?

No. It measures demand. Ask about safety, care, and whether everyone leaves with a next step.

What share of people experiencing homelessness in Canada are Indigenous?

33% of respondents in the 2024 national count, compared to about 5% of the Canadian population.

Is homelessness getting worse in Canada?

Yes. Analysis published by the Canadian Alliance to End Homelessness in 2025 puts the increase at roughly 11% between 2018 and the 2020 to 2022 counts, and a further 79% between those counts and 2024. The same analysis argues the driver is income after rent, which is outside any single charity's control.

What is a By-Name list?

Real-time data on every known person experiencing homelessness in a community, shared between agencies so support is prioritized by need.