How to evaluate a real estate CRM: the twelve criteria

We sell a CRM, so this post has an obvious conflict of interest and I resolve it the only way that seems honest: the criteria work even if you do not choose us, there are no brands and no third-party prices — those change and quoting them wrong would be unfair — and our own answers go at the end, in a clearly marked section of their own.

The reason to take this seriously is simple: a CRM gets replaced every few years and the real cost is not the subscription, it is the migration, retraining the team and the months when nobody uses it properly. Choosing carefully the first time is far cheaper than choosing fast.

The twelve criteria

• 1. Data ownership. Do the leads belong to the brokerage or to the agent who worked them? If an agent leaves, what goes with them? Get it in writing before you sign; it is the question that costs the most to learn late.
• 2. Full export, without asking permission. That you can download contacts, notes, history and documents whenever you want, in an open format and without opening a ticket. A product that makes leaving difficult is compensating for something.
• 3. Migration in. How your current data gets in, which fields map, and what happens to the history. A half-done migration is worse than starting clean, because the team stops trusting what it sees.
• 4. Genuinely bilingual. Not a translate button: the whole interface in both languages, templates in both, and the same file worked by two people in different languages. In Florida this decides hires.
• 5. Full mobile, not a viewer. The test is to do everything an agent does in a day from the phone, without opening a laptop: call, log, move a stage, send a document. If halfway through you have to "go to the computer," it is not mobile.
• 6. Native commissions. That commission, splits and overrides are calculated inside and export to accounting. The alternative is a parallel spreadsheet, and the parallel spreadsheet always ends up being the truth, which makes the CRM decorative.
• 7. Automation you can understand. Being able to see which automations are live, what triggers each one, and switch them off without calling support. An automation nobody can explain is a time bomb with your clients inside.
• 8. Integrated communications. Calls, messages and email logged in the file without copy-paste. If the conversation lives on the agent’s personal phone, the brokerage does not have a history: it has rumors.
• 9. Reports for the broker, not just the agent. Funnel by stage, activity per person, production, and a decent export. The seven numbers that actually decide something are in another guide; what matters here is whether the system hands them to you unprompted.
• 10. External roles. If the lender, the title attorney or the coordinator have to log in, ask whether they occupy a paid seat and what exactly they can see. That is a cost and privacy difference, not a technical detail.
• 11. Total price, seats and extras included. Count everyone who logs in to work (the broker included) and add messaging, calls, storage and e-signature if they are billed separately, and ask for the full annual figure. The number in the ad is never the number on the invoice.
• 12. What happens if you leave. How much notice is required, whether the account goes read-only or is deleted, how long you have to export. Asking before you sign tells you more about a vendor than the entire demo.

The eight demo questions

A demo done right is not run by the salesperson: it is run by you. Eight questions, and you write the answers down.

• "Show me the screen an agent sees on Monday morning." If they struggle to find it, that screen does not exist.
• "If I want to leave tomorrow, how do I get my data out and how long does it take?"
• "Show me the same file in English and in Spanish."
• "Do all of it from the phone, without touching the computer."
• "How is a commission calculated with two agents and a referral?"
• "What happens when an agent leaves: what goes with them and what stays?"
• "What counts as a seat, and what would a full year cost with my current team?"
• "Show me something the product does not do yet." The answer to this one tells you who you are talking to.

Warning signs

• The demo only visits marketing screens and never the boring work of an ordinary Tuesday.
• Nobody can tell you what the product does NOT do.
• Migration is "something we look at after signing."
• The price depends on who you talk to.
• You have to call support to switch off an automation.
• They promise the AI predicts who is going to sell. There is a whole guide about that.

How to test it in fourteen days, with method

A trial without a plan always ends the same way: opened on day one, looks nice, decided on instinct. This is the plan we would use to test any product, ours included.

• Days 1-2: import real data, not samples. Fifty actual contacts, with their duplicates and their empty fields.
• Days 3-5: two agents work their real leads there and only there. No dual system: the test breaks the moment you can fall back.
• Days 6-8: one complete file from start to finish, documents and a signature included if there is one.
• Days 9-11: the broker pulls their numbers with no help from anyone at the vendor.
• Days 12-13: language and phone test, with someone who actually works in Spanish.
• Day 14: export everything. If this goes badly, it does not matter how well the rest went.

Who decides, and why that decides the outcome

The CRM is almost always chosen by the broker and used by the agents, and that is where half of adoption failures begin. The tool can be excellent and still die in three months if the team experiences it as something dropped on them. The fix is not a vote: it is putting two people into the trial from day one — the highest-volume agent and the most skeptical one. The first finds the real limits; the second finds the excuses that later turn into resistance.

And one question worth answering before signing rather than after: what happens to whoever does not use it. If the answer is "nothing," the system will coexist with three parallel spreadsheets and the brokerage will not have data, it will have versions. A CRM only works when it is the single place where the truth lives, and that is a management decision, not a product feature.

How we answer the twelve

This is the interested part, which is why it sits on its own. Leads belong to the brokerage from day one: if an agent leaves, they stay. Data is always exportable. Data comes in by CSV, free and self-service, with assisted or full migration if you would rather we did it; the detail is in migrate your leads and the prices are on the pricing page. Our real estate CRM for brokerages is bilingual across the whole interface, not through a translate button. Commissions are calculated inside and export to accounting. Closing professionals — lender, title, coordinator — do not take a seat in your brokerage: each one pays their own plan. And on seats we say it where it shows: the broker counts as one more.

What we do not do yet counts too, and we keep it written on the product pages rather than buried in a footnote. If you evaluate any CRM (ours included) with the twelve criteria above and the eight questions, you will make a better decision than almost anyone does. And if that leads you to someone else, honestly, we prefer that to having you unhappy for two years.

If you want to see the full flow before talking to anyone, the Academy course on agency CRM walks through it step by step.