My marketplace has 340 sellers and 11 buyers, and the buyers are the problem

$31,000 of ad spend for 340 submissions is $91 per submission, and 9 closes means $3,444 in acquisition cost per closed deal against a $2,220 average fee. You're not underwater on timing, you're underwater on unit economics as currently structured.

To get to breakeven at 1.5% you need either the close rate up from 15% to about 24%, or acquisition cost down to roughly $55 a submission, or the fee up. Those are the three levers and only one of them is under your control this quarter.

Which one do you think moves? Because if the answer is the fee, the buyer retainer idea is really just a fee increase with better collection mechanics, and you should price it that way.

Two buyers going around you after six months means eleven will do it after eighteen. The behavior spreads because they talk to each other.

The only marketplace defense I've seen work is that the platform holds something they need. Escrow coordination, the title order, whatever it is in your state. If the deal can complete without you touching it, it will.

Asking the thing I'd be embarrassed to ask: how does a seller who submitted to your site even end up in contact with a buyer directly? Is the buyer getting the address and the phone number the moment the listing goes live?

If so, that seems like the leak, and it's a product decision rather than an enforcement problem.

compass identified the structural issue. Leakage in a lead marketplace is a function of when you release identifying information, and most platforms in this space release it too early because early release is what makes buyers happy.

The mechanism that reduces it: stage the disclosure. Address block and property detail visible to all buyers, exact address and seller contact released only after the buyer commits to something, either a signed access agreement, a deposit, or acceptance of the fee obligation on any transaction with that seller for a defined tail period. The tail period clause is standard in brokered transactions and it's what makes the obligation survive the buyer walking away and coming back in ninety days. It only helps if it's enforceable, and enforceability of a tail on a non-licensed intermediary varies by state, so that clause needs to be drafted by an attorney licensed where you operate. You already have that call booked.

On the buyer concentration: four buyers at 80% is not unusual at your stage and it's also the reason your take rate is under pressure. A retainer changes the negotiation because the buyer has paid for access before they see a deal, which raises the cost of going around you from zero to whatever they already paid.

The assumption your whole model rests on is that the automated range can be accurate enough to produce a live listing buyers trust. At 15% accept-to-close, it currently can't. Every other fix is downstream of that. I'd hold the buyer restructure until you know whether the range can be made accurate, because a retainer sold on top of listings buyers don't trust churns in one cycle.

How did the seller describe the direct approach? Did the buyer say something like "we can do this without the fee and I'll pay you more"? Because if the buyer is splitting the saved fee with the seller, both sides are motivated against you and terms won't hold.

The retainer version is the only one I'd fund, and only if the retainer covers your ad spend. Four buyers at $2,500 a month is $120k a year, which is most of your acquisition budget, and then every close fee is margin.

But that only works if four buyers believe the flow is worth $30k a year each. Nine closings across everybody in six months means the average buyer got two deals. Would they pay $30k for two deals? Depends entirely on what those deals made them, and you probably don't know.

340 submissions and only 61 accepted the range. That's 279 sellers who raised their hand and then walked. Isn't that the bigger number in the post? Feels like there's a business in the people who said no.

ember has a point. Those 279 didn't say they won't sell, they said they won't sell at that price. Some percentage of them list with an agent within the year and some just wait. A follow-up sequence against that group costs almost nothing compared to $91 a submission.

In the rural markets I look at, an automated range would be hopeless. Three comps within five miles and two of them sold in 2021. harrow, are your 340 concentrated in one metro or spread out? Because the fix for the range problem is completely different depending on comp density.

The two buyers going direct are telling you the fee is too high relative to what they think you did. 1.5% on $148k is $2,220 and in their head they did the negotiation and the diligence and you sent an email.

Doesn't mean lower the fee. It means the platform needs to visibly do more of the work. If you're coordinating the title order and chasing the seller for documents, $2,220 stops looking like a toll.

Following up on my own question because harrow hasn't answered and I think it matters more the longer I sit with it. If the 340 are spread across a wide area, your buyer pool of 11 is thin in most places by definition, and the four active ones are probably all working one submarket. That would explain the concentration without anything being wrong with the marketplace design.

In that case widening the buyer pool isn't dilution, it's coverage, and the leakage worry is a separate fight.

Running the mail version of this on a much smaller scale and the thing I'd take from your post is the 15%. I'd stop calling it a range problem until you've called the sellers who accepted and then didn't close and asked them what happened. Nine closings out of 61 live listings means 52 conversations available to you right now, free.

My guess is you find three or four distinct failure modes and only one of them is pricing. Title issues, a spouse who wasn't on board, a buyer who inspected and retraded. Each of those has a different fix and sextant's levers change depending on which one dominates.

The retainer decision can wait two weeks. Those calls can't, they get colder every day.