38,000 owner records and two paying seats. How do I sell the third?

Run agency, sell things that look like this. The churn number you're missing is the one that kills option one, and I can tell you roughly where it lands. On a data-access product with no done-for-you component, expect the median subscriber to cancel between month two and month four. They pull the list, get intimidated by the mail cost, and blame the data.

The two people paying you at month five are not representative. They're friends of yours who feel awkward canceling.

If you want seats to stick, the seat has to include the mail going out. Charge $349 plus per-piece and press send for them. Retention on that is a different animal because they're seeing pieces land and calls come in, and the cancel decision costs them momentum instead of saving them $349.

What counts as a lead in option three? You said live callback with stated motivation, which is doing a lot of work. Is your VA qualifying for price expectation and timeline, or just confirming the person is the owner and picked up the phone?

Because $140 for a warm human who said they might sell in a year is a bad buy, and $140 for someone who named a number is a steal. Same label, different products.

The assumption your three options hang on is response decay per additional subscriber in the same territory, and you can measure it before you pick.

Your current 4.1% response is on records nobody else is mailing. Add a second mailer to the same 38,000 and the marginal effect isn't a clean halving, it depends on overlap in filters. If your subscribers use the same tax delinquent plus long tenure filters you tuned, overlap will be high, probably 60% or more of pulled records, and both of you see response drop toward the 2% range on the overlapped set. If they filter differently, decay is much smaller. Run one month where you give a subscriber a filter set you deliberately don't mail and compare.

On the economics: at three contracts a quarter your all-in cost per contract is about $3,900. That's the ceiling on what a per-lead price can be, because a rational buyer prices against building it themselves. At $140 a lead with a 1-in-90 conversion, a buyer pays $12,600 per contract, which only works if their spread per deal is large and their alternative is nothing.

Exclusivity at $1,900 prices the territory at roughly half your delivered cost of running it, which is defensible if the subscriber replaces your mailing entirely and you exit the county.

The practical sequence is to sell seat three non-exclusive at a higher price than $349, instrument the response decay, and only commit to a model in month four when you have cohort data. Pricing decided before you have decay numbers is a guess.

Data reseller agreements often restrict redistribution, so confirm your bulk provider's terms in writing before you sell access to anyone.

Nobody's said support yet. Two subscribers means two emails a week. Twenty means a person, and that person costs more than the $349 seats bring in until you're at maybe fifty.

I run property management software conversations all day and the ratio that matters is tickets per account per month. Anything above 1.5 and self-serve is a lie you're telling yourself.

cobalt is right about the friends. I'd add that option one doesn't just dilute the leads, it makes your own mail worse and you're the one with three contracts a quarter riding on it. You'd be selling the thing that pays you.

Exclusivity is the only version where your interests and the customer's point the same way. $1,900 feels low to me though. If a county produces three contracts a quarter for you and a contract is worth what I assume it's worth, the exclusive holder is buying $40k+ of annual deal flow for $22,800. Price it at what it produces.

Where does the $900 bulk data come from and does the license let you resell derived records? I've been pricing the same thing and every provider I called had a different answer about what "internal use" means.

Do the cohort math before you sell anything else. If seats churn at 20% monthly, a $349 seat is worth about $1,745 in lifetime revenue. If they churn at 5%, it's $6,980. Same price, completely different business, and one of those supports a sales effort and the other doesn't.

You can't know which one you have from two friends at month five. thicket you said it yourself. So the next three sales should be to strangers, deliberately, even if you have to price them lower to get them.

I'm the customer you're describing and I've canceled two of these. Both times in month three, both times for the same reason: the list was fine and I never sent the mail.

What would have kept me is if the tool had told me I was behind. Some kind of "you pulled 1,200 records and mailed 0" on the dashboard. Sounds trivial. It's the difference between me feeling like I'm failing and me feeling like the product is.

Disagree with cobalt on bundling the mail. The moment you press send for people you're a mail house with a database attached, and mail houses have terrible margins and a whole operational tail. Reprints, address errors, the client who wants a different letter.

thicket already has a working funnel. Adding fulfillment for other people's campaigns is a second business, not an upsell.

Your filter set is aimed at operators buying 1-4 unit and small multifamily. Land people would pay for the same infrastructure with completely different filters, and there are fewer vendors chasing them. Tax delinquent on vacant parcels, out of state owner, no improvements on the assessment. Might be worth pulling one county's land records just to see the count.

What happens to the two paying seats if you go exclusive on their county?

Dumb question, why does the same record going to more people make the response rate worse? Isn't it the same owner either way?

rigging, the owner gets four letters in a month instead of one and stops reading any of them. And the ones who do call have four offers, so your conversion drops on top of your response.

thicket, before any of the pricing, I'd want to see your data provider's contract. Specifically what counts as a derived work. If you enrich a record with a skip trace and a score you built, some agreements still treat the underlying record as theirs and prohibit passing it on. That clause has ended products like this before they had customers.

The second lien flag is the most interesting thing in your post and you spent one line on it. That's a distress signal most list vendors don't surface at all.

I'd want to know your response rate on records with that flag versus without. If it's meaningfully higher, that's the product, and you sell 4,000 scored records instead of 38,000 raw ones. Small, expensive, defensible.

Sell seat three at $349 to a stranger and see if they're still there in March. Everything else is a decision you can make later.

How are you handling the phone side? If the VA is calling out on skip traced numbers rather than only taking callbacks, the rules around calling cell phones and scrubbing against do-not-call lists apply to whoever places the call, and once you're providing numbers to subscribers that gets more complicated. This is an area where the exposure is real and the rules change, so I'd have an attorney who handles this specifically look at how you structure the seat before you sell access to phone numbers.

Your fixed cost is $2,640 a month and you're running three contracts a quarter. So the internal business is fine on its own, assuming a contract is worth more than $880 to you, which I'll assume it is or you'd have stopped.

The question I'd ask is what the platform business needs to be worth for it to justify what it costs you. Not in dollars, in attention. Twenty subscribers is a support load, a billing system, refund arguments, and a roadmap. Meanwhile the three contracts a quarter are being generated by you, personally, and you're the one who'd stop doing that work.

I'd want the platform revenue target to be at least 3x what the internal funnel nets before I'd trade one for the other. If $5,700 exclusive is the ceiling, the answer is probably that this is a good private list and not a business.

Long ownership tenure is the filter that ages best. Tax delinquency comes and goes with a refinance. Somebody who's held 12 years and lives three states away is still going to be tired in 2027.

anchor named the right document. I'd add the skip trace vendor's terms, which usually have their own restriction on passing results to third parties, and often a separate clause about using the data for marketing at all.

So you have two upstream agreements that both need to permit what you're planning, and neither was written with your business model in mind. Read both, then have someone licensed read the two clauses you're least sure about. Most of these products get built and then discover the license problem when they have paying customers and can't unwind it.

When you say one in 90 responses becomes a contract, is a response someone calling back, or does a returned letter count? Trying to compare this against numbers I've seen elsewhere and I don't think everyone means the same thing.

juniper, that's the right question and it's why published response rates are useless for comparison.

thicket, if you go per-lead you need a written definition of a qualifying lead and a replacement policy in the terms before the first invoice. Wrong number, deceased owner, already sold, owner says never contact me again. Decide now which of those you replace and which you don't, because the argument in month two is worse than the concession.

Your second lien flag depends on recorded documents, and how easy those are to get in bulk varies enormously by state and even by county. Some recorders sell a monthly bulk file, some make you scrape an image viewer, some charge per document. Also worth knowing that roughly a dozen states don't make sale prices public at all, so if you ever want a sale price field in the scoring, your coverage map has holes in it that no vendor can fully fix.

That matters for expansion pricing. Three counties you can maintain by hand. Thirty means the acquisition cost per county is wildly uneven and your margin story changes.

I'd want to see the two subscribers' actual usage, not their payment status. How many records did each pull last month, and did either mail anything? If they're paying and dormant you have two months of runway on that revenue, maximum.

Usage data is the only leading indicator of churn you have at this size.

tally's exclusive math is right but I'd take it further. Don't sell access at all, sell a cut. Give the county to one operator for a low base and a fee on every contract that comes out of your leads. You're already tracking which callback became which contract or you wouldn't know your 1-in-90.

Collection is the hard part and I've been burned on a similar arrangement, so it needs to be documented properly and you need an attorney to structure it. But it aligns better than any flat number and your upside isn't capped at $5,700.

flint, the problem with a fee on contracts is that depending on your state, taking compensation tied to a real estate transaction you helped bring about can look like brokerage activity and trigger a licensing question. Whether it does depends on how the fee is structured and which state you're in, so that's a conversation with a licensed attorney in the state before you write it, not after.

A flat subscription doesn't raise that question in the same way, which is part of why everyone in this space sells subscriptions.

Reading this whole thread as someone still choosing a lane, the thing that stands out is that thicket has a working acquisition funnel and is considering trading it for a software business he hasn't validated. Every reply here is essentially saying the funnel is the asset.

Not sure that's the answer, but it's a lot of independent people arriving at it.

I followed about 70% of this. The part I'd still want spelled out is what the two current subscribers actually do with the login, because canton asked and thicket hasn't said, and it seems like the whole thread turns on it.