Your CRM doesn't know what a transaction is

Pull up a contact in almost any real estate CRM and look at what it knows.

Name, contact details, source, some tags. A timeline of emails and calls. A stage on a pipeline. Maybe a lead score, maybe some notes from a showing eight months ago.

Now ask it when the inspection contingency expires.

It has no idea. It doesn’t know there’s a deal. As far as the CRM is concerned, this person moved to a stage called “Under Contract” and then, weeks later, to one called “Closed,” and everything that happened in between took place somewhere else.

Why are CRMs and transaction platforms separate?

Not because anyone decided they should be. It’s an accident of when these products were built and who bought them.

CRMs were sold to agents to solve lead follow-up. That was the problem fifteen years ago and it’s still a real one. Transaction management was sold to brokerages to solve compliance. Brokers carry the liability for file completeness, so brokers bought the software, and it was designed around the review queue rather than around the agent.

You can still see it in how the products describe themselves. SkySlope’s homepage, for example, calls it “the AI-first way to run a brokerage,” with “proactive compliance” and “automated auditing.” That’s a brokerage product, and a good one for that job.

Two different buyers, two different problems, two categories of product. The agent ended up with both, and with the job of joining them. That made sense when both were simple. It makes less sense now that a CRM costs $69 a user and a transaction platform about another $35, and neither one can answer a question about the other.

Where does each piece of a deal live today?

In a split setup, in at least three places. Here’s a typical deal traced from first inquiry to the call eighteen months later.

Every row where the two columns differ is a place something can go missing, and a place you type the same information twice.

What does the gap actually cost?

Three things, in order of how much they hurt.

Deadlines live outside your task list. This is the expensive one. Your inspection deadline, your financing contingency, your appraisal window: they’re in the transaction platform, which emails you about them. That email goes to the same inbox as everything else. Meanwhile the task list you actually work from is in the CRM, and it has no idea any of this exists. Most agents catch these anyway, through diligence and a good calendar habit. But “most” and “anyway” are doing heavy lifting in that sentence.

The past client is a stranger again. Years after closing, they call about selling. The relationship is in the CRM. The transaction (the property, the numbers, the problems you solved) is in a system you may not even pay for anymore. So you go looking.

Nobody can tell you how long anything takes. Average days from contract to close. Which lender consistently runs late. Whether your file completion actually improved after you changed your process. All of these need transaction data joined to contact data. If those live in two systems with no shared key, the analysis just doesn’t happen.

Which dates does a deal actually run on?

The ones in the contract, plus a few set by law. The exact periods come from your state’s contract forms and what the parties negotiate, so treat this as the shape rather than the numbers.

Notice that “N days” deadlines depend on counting correctly: calendar days or business days, and from which date. That’s easy to get wrong by hand at 9pm, and it’s the kind of thing software should do.

What should a combined system do?

Treat the deal as part of the client, and turn the contract’s dates into work. Here’s the checklist we’d hold any product to, ours included.

• The deal hangs off the contact. Open the client, see the deal. Open the deal, see the client’s whole history.
• Deadlines become tasks on the same list as your follow-up calls, not reminders in a separate inbox.
• Dates come from the contract, not from retyping. Reading the dates out of the signed contract, and counting business days correctly, removes the most common error.
• Signed documents file themselves against the deal they belong to, with the signing history attached.
• Something checks the file. Missing documents, unsigned forms and deadlines due this week should surface on their own.
• The deal stays after closing, so the next conversation starts with the full story.
• Everything exports. Deals, documents and certificates should leave with you if you ever move.

In Workspace, the transactions hub is free on every plan: deals, milestones, close dates, documents and e-signature, and it all stays with the client record. Deal deadlines show up on the same Today list as your calls and replies, and deal dates are copied into your Google Calendar alongside your appointments. On every plan (on Free, from its monthly taste of AI), the assistant reads a contract once and fills in the milestones, such as earnest money, inspection, appraisal, financing and closing, counting “N days” deadlines on the server. It never overwrites what you typed, and it flags a price that doesn’t match. File review then points out checklist gaps, missing documents, documents still waiting on signatures and milestones due within three days.

What happens to the file after closing?

It has to be kept, and it has to stay useful. Those are two different requirements.

Kept, because the law says so. Record-keeping rules are set by each state and apply to the broker. In California, for example, a broker must retain copies of listings, deposit receipts, canceled checks, trust records and other transaction documents for three years. Other states set their own periods, and your broker’s policy may be longer. None of that changes where your copy should live.

Useful, because the client comes back. NAR’s 2025 Profile of Home Buyers and Sellers found sellers had stayed in their homes a median of 11 years, an all-time high. A record that has to be useful a decade later can’t depend on still paying for the system it was created in. That’s the strongest argument for keeping the deal with the client, and for being able to export it.

When does a separate compliance platform still make sense?

When your brokerage requires one, or when compliance at scale is the job. There’s a real counterargument here, and it’s worth stating.

Brokerage compliance is its own job. Broker review queues, state-specific checklists and audit trails across hundreds of files a year are what platforms like SkySlope and dotloop are built around. Workspace is built around the agent’s side of the deal: the client, the dates, the documents and the signatures, on one record.

There’s also a network effect in transaction software. dotloop is easier when the agent on the other side of the deal is already on dotloop. That’s genuinely valuable, and it’s the strongest argument for staying put.

So the honest position isn’t “transaction platforms are obsolete.” It’s that the split was never designed, and most agents inherited it rather than choosing it. If your broker requires a compliance platform, keep using it for file review, and keep the deal, its dates and its documents with the client too. Our guide to transaction management software compares the options, including how Workspace and dotloop differ.

How do you test your own setup?

Take your three most recent closings. For each one, count how many separate systems you’d have to open to reconstruct the full story: first contact, showings, offer, contract, deadlines, documents, closing.

If the answer is one, you’re in good shape and you can ignore all of this.

If it’s three or four, that’s the shape of the thing. Whether it’s worth changing depends on how often you actually need that story, which, if you do any repeat or referral business at all, is more often than you’d think.