How real estate agents actually use vendor referrals: a field guide

Every real estate transaction generates a small economy of secondary referrals. The buyer needs a mover, a cleaner, sometimes a painter or handyman before they unpack. The seller needs a stager, sometimes a landscaper, often a junk removal crew. Each of these jobs is worth real money to the vendor who gets it. None of it is captured by the MLS or the commission check.

What’s interesting isn’t that these referrals exist. It’s that they happen entirely outside any system. There’s no platform, no record, no formal compensation. The agent texts a vendor’s number to a client. The vendor calls the client. Work gets done. The agent might get a thank-you card.

I’ve spent the last two years building infrastructure for this informal economy, and along the way I’ve talked to about 40 real estate agents across Atlanta, Nashville, and Charlotte about how they actually do this work. Here’s the field guide that emerged.

The four referral patterns

Agents don’t all refer the same way. There are roughly four patterns:

Pattern 1: The text-thread agent

Most common. The agent has 8 to 15 vendors in their phone contacts, organized loosely by trade. When a client mentions needing a painter, the agent texts the painter’s number along with a brief note like “tell him you got referred by [Agent], he’ll take care of you.” That’s the entire transaction.

The advantage of this pattern is speed and intimacy. The disadvantage is that nothing gets captured. The agent has no idea if the client ever called the painter, whether the work happened, whether anyone was happy. The vendor doesn’t know which agents are sending real business and which are sending tire-kickers.

This is probably 70% of all agent referrals today.

Pattern 2: The Notion-doc agent

About 15% of agents have built some version of a curated vendor doc, usually in Notion, Google Docs, or a vintage WordPress site they maintain themselves. They link to it from their email signature, share it at closing, sometimes print PDF versions.

This pattern produces measurable value for the vendor (a few documents I’ve seen get 50-100 page views a month) but it requires the agent to maintain it and never bills the vendor for being on it. It’s the closest thing to ClosingPack that exists today, just unmonetized.

Pattern 3: The brokerage-program agent

Larger brokerages (Compass, Coldwell Banker’s better franchises, some independents) have started building “preferred vendor” programs at the brokerage level. The brokerage negotiates blanket relationships with cleaning companies, moving companies, etc., and individual agents are nominally part of them.

These programs almost never produce good outcomes for the working agent. The brokerage captures whatever value exists. The agent loses the ability to recommend their own preferred vendors. The vendor relationship is national or regional rather than personal. Almost every agent I talked to who’s part of a brokerage program also maintains a Pattern 1 or Pattern 2 list because the official program doesn’t work.

Pattern 4: The high-volume team

A small number of agents (especially team leads on production teams closing 80+ homes a year) have built actual systems. Often a transaction coordinator maintains the vendor list and sends introductions on behalf of the team. Sometimes vendors pay an annual sponsorship to the team. These arrangements work but they’re expensive to operate and only feasible at very high volume.

This is maybe 3% of the agent population. They’re closest to what ClosingPack offers, but most of them are spending thousands of dollars a year operating it themselves.

Where the value actually gets created

Step back from how it’s organized and look at what’s actually happening. There are three parties exchanging value:

The homeowner gets a vetted vendor recommendation at a moment when they’re spending more money than they’ve ever spent in their life and have low context on local trades.

The vendor gets a warm referral that closes at roughly 25%, compared to about 3% for cold leads from directories like Angi.

The agent gets the satisfaction of helping their client, plus a small reputational benefit if everything goes well.

Notice what’s missing. The agent, who’s doing the actual matchmaking, is the only one not getting paid for the value they create. The vendor walks away with revenue. The homeowner walks away with a working vendor. The agent walks away with maybe a thank-you text.

This is the part of the system that’s been busted for decades. It’s not that referrals don’t have value. They have a lot of value. It’s that there hasn’t been infrastructure for the agent to share in it.

Why nobody fixed this earlier

A few reasons:

Legal confusion. Most agents I talked to assumed accepting payment for vendor referrals was illegal. It isn’t, with the structural and disclosure conditions covered in the RESPA explainer. But the confusion has kept agents from formalizing.

No platform. Building a vendor list, accepting recurring monthly payments from vendors, paying out to agents, handling disclosures, doing tax reporting, calculating splits, none of this is something an individual agent can build. It requires platform infrastructure.

Cultural resistance. Some agents recoil at the idea of “monetizing” client relationships. This is real and worth taking seriously. The good versions of vendor referral monetization don’t change anything the agent would do anyway. They just attach a small payment stream to the recommendations that were already happening for free.

Fragmentation on the vendor side. Home service vendors are an extraordinarily fragmented market. There are 4 million small home service businesses in the US. No platform has ever managed to organize them at the city level in a way that holds up.

What changes with formal infrastructure

When you put real infrastructure underneath this informal economy, three things change.

Better vendors get more work

When vendors pay a monthly fee for a placement, they’re investing in being on the list. That investment changes their behavior. They follow up faster on referrals. They prioritize agent-referred customers. They send the agent a heads-up when a job goes well or badly. The relationship becomes professional rather than passive.

This is the most underrated effect. The same vendor who used to ignore half their referrals from text-thread agents will treat agent-referred homeowners with the seriousness they deserve when there’s actual money on the line.

Worse vendors lose their seat

Three vendors per category, per agent. That’s the ClosingPack constraint. It means every slot is earned. The vendors who don’t deliver get cancelled by the agent. The vendors who do deliver keep their slot for years.

Compare this to Angi or Thumbtack, where any vendor with a credit card can buy as many leads as they want. The platform incentive is volume, not quality. Bad vendors stay in the pool indefinitely because they pay per lead.

ClosingPack’s incentive is the opposite. The agent’s reputation is on the line for every recommendation. The bar is higher. Vendors who clear it get a real business outcome.

Agents capture some of the value they create

This is the obvious change but it’s also the most controversial. Some readers will see this and immediately think “kickback.” That word is doing a lot of work.

The legal framework here is clear and the structural design (flat monthly fees, non-settlement services, disclosed to homeowner) puts ClosingPack outside RESPA’s prohibition. But the cultural objection is separate. Is it weird for an agent to be compensated for vendor recommendations they would have made for free?

My honest answer: it shouldn’t be. Magazines accept payment from chefs to feature their restaurants. Newspapers run sponsored content boxes. Newsletter writers get paid by tools they recommend. Real estate agents are the only category of professional curator who haven’t been paid for their curation.

The “kickback” framing assumes the agent is selling out their client. Done correctly, the opposite is true. Formalizing the relationship raises vendor quality, which serves the client better. The agent earns a small recurring fee. Everyone wins except the categories that depend on the agent’s free recommendations being uncontested.

What this means for vendors

If you’re a home service vendor reading this, here’s the practical implication: in 5 years, every working real estate agent will have some version of a vendor pack. The question isn’t whether this happens. The question is whether you’re on the right agents’ packs in your market.

Today there’s no friction to being on them. ClosingPack offers 45 days free, with a flat monthly fee thereafter. The slot is yours until you cancel. You don’t need to be a national brand. You need to do good work for a few homeowners and let the agent vouch for that work.

Five years from now, the slots will be valuable enough that vendors will compete for them. The early ones are the cheap ones.

What this means for agents

Most agents I’ve talked to vastly underestimate the volume of vendor referrals they’re already making. Try the exercise yourself: open your text messages from the last six months and count how many times you’ve sent a vendor’s number to a client.

The number is usually between 30 and 60. For active agents, often more than 100.

Each one of those was a value-creating act for the vendor. Today you’re doing it for free. There’s no requirement that you keep doing it for free.

ClosingPack exists to let agents capture some of that value with clean infrastructure, RESPA-compliant structure, and full disclosure to the homeowner. None of it changes the recommendations you’d make anyway. It just attaches a small recurring revenue stream to them.

That’s the field guide. The next 10 years of this market belong to the agents who realize the recommendations they’ve been giving away for a decade have always been worth money.