A firm in Ohio wrote to us last spring, annoyed. They had upgraded to Premium two days earlier, paid the annual amount without complaint, and expected a badge to appear on their profile by morning. It did not. The badge slot on their page still read unverified. Their reasoning was direct: I paid, so where is my checkmark.<\/p>\n
We answered the way we always do. Payment moved them up the placement rows. It gave them a dofollow link once their expanded profile was published. It did nothing to the badge, and it never will, because the badge does not read from the billing table. It reads from a different table entirely, one that only fills when a human on our side has looked at documents and recorded what those documents showed.<\/p>\n
That exchange is the whole argument of this post compressed into three emails. Most legal directories collapse those two things, payment and trust, into one purchase. We split them on purpose, at the level of the database, and we want to explain exactly how the split works, what each side of it can and cannot do, and why we think the seams between them are the most honest part of the product.<\/p>\n
Start with the money, because that is the part everyone understands.<\/p>\n
We sell placement and visibility. We are not shy about that. A firm that pays for Premium gets several concrete things, and each one is a feature, not a judgment about the quality of their lawyering.<\/p>\n
They get premium placement rows. On a city or practice-area listing page, paid tiers sort above free profiles. That is a paid position, and we label the ordering logic openly so nobody has to guess whether the top firm is there because it is best or because it bought the row. More on the exact ordering below, because the detail matters.<\/p>\n
They get an expanded profile. A free listing carries a name, a location, a practice area, and, if the firm has earned it, a verification badge. A paid profile carries more: a longer description, additional practice areas, attorney bios, office photos, a contact form that routes to the firm directly. These are visibility features. They help a firm tell its story. They do not certify that the story is true.<\/p>\n
They get a dofollow link, but only under a condition that trips a lot of people up. The link is dofollow when the profile is both paid and published. Payment alone does not publish anything. A firm can pay on a Tuesday and, if their submitted content is still in the review queue, sit unpublished on Wednesday with no live link at all. The dofollow attribute attaches to a published page, and publication is a separate gate operated by editors, not by the payment processor.<\/p>\n
That last point is where our structure starts to diverge from the industry norm, so let us slow down on it.<\/p>\n
Payment never auto-publishes. When a firm completes a transaction, our system records the transaction and unlocks the ability to submit expanded content. It queues that content. It does not push it live. A person reviews the submission for the basics: is this a real firm, does the described practice match public records, are the attorney names attached to real bar registrations, is anything in the text defamatory or fraudulent on its face. Only after that review does the page publish, and only after it publishes does the dofollow link exist.<\/p>\n
We hold the same line against automation from the other direction. Nothing scraped auto-publishes. Nothing AI-generated auto-publishes. We ingest a lot of public data to seed profiles, bar directories, court records, firm websites, and it would be trivial to let that pipeline write pages and push them live at machine speed. Plenty of directories do exactly that, and you can tell, because their pages read like they were assembled by something that has never spoken to a lawyer. We stop the pipeline before the publish step. A draft assembled from scraped fields sits as a draft until a human clears it. The same is true for any text a language model helps draft internally. The model can propose. It cannot publish.<\/p>\n
So the honest summary of a tier is this. You are buying reach and presentation. You are buying the row, the expanded page, and, once an editor publishes it, the link. You are buying our audience’s attention. You are not buying our audience’s belief, because belief on this directory is minted somewhere the payment system cannot reach.<\/p>\n
Here is the list of things no amount of money moves. We keep it short deliberately, because a short list is easier to hold ourselves to.<\/p>\n
Money cannot buy a verification badge. The badge is not a graphic we drop onto paying customers. It renders from a stored verification row, a record that says which credential was checked, on what date, against what source, by whom, and with what result. If that row does not exist, the badge does not render, and there is no manual override button in the billing interface that summons one. We built the display logic so that the badge template queries the verification table and only the verification table. A Premium subscription with an empty verification record shows no badge. A free listing with a full verification record shows the badge. The two are wired to different power sources.<\/p>\n
Money cannot buy a passed check. When our reviewers work through a verification, they can record a pass, a fail, or a pending state on each item. Bar standing is one item. Firm identity is another. Claimed practice areas, office existence, and disciplinary history are others. A firm that pays us the largest sum on our price sheet and fails the bar-standing check gets a recorded fail. The fail sits in the row. We do not delete inconvenient fails for good customers, because a verification record that can be edited by the accounts-receivable department is not a verification record, it is a receipt wearing a costume.<\/p>\n
Money cannot buy publication. We said this above, and we are saying it again because it is the load-bearing wall. Publication is an editorial act. It happens after review. A payment queues content for that review and nothing more. We have unpublished pages belonging to paying firms. We have published pages belonging to firms that have never paid us a cent. The two facts coexist every day in our database, and their coexistence is the point.<\/p>\n
People sometimes ask whether these are really separate, or whether we just wait a polite interval after payment and then rubber-stamp everything so the separation is theater. Fair question. The answer is in the fail rate. If payment quietly guaranteed a pass, our fail column would be empty for paying firms, and it is not. We record fails against paying accounts. We record pending states that outlast the billing cycle because a firm never sent the document we asked for. A subscription that renews does not retroactively resolve a verification item that was never satisfied.<\/p>\n
Ordering is where directories hide their real incentives, so we will put ours in the open where you can check them.<\/p>\n
The sort runs in a fixed sequence. First, tier. Paid tiers sort above free listings. This is the part we are paid for, and we are not going to pretend otherwise by burying paid firms and calling ourselves pure. If you pay, you rank higher on the page. That is the deal, and it is disclosed on the page itself.<\/p>\n
Second, within a tier, validated-review rating. Reviews on this directory are validated before they count toward the rating, which means we take steps to confirm a review reflects a real client interaction rather than a competitor, a bot, or the firm’s own marketing intern. Unvalidated reviews can appear, but they do not move the sort. So within the paid tier, a firm with a higher validated rating sorts above a firm with a lower one. Within the free tier, the same rule applies. Rating orders firms inside their tier band, never across it. A brilliant free firm with a perfect validated rating still sorts below a mediocre paid firm on the placement page, and we tell you that outright rather than letting you assume the top row is a merit ranking.<\/p>\n
Third, a deterministic tie-break. When tier and rating are equal, we need something to decide order, and we refuse to make it random or make it purchasable. We use a deterministic key, the same input producing the same output every time, so that two firms with identical tier and identical rating land in a stable, reproducible order that nobody bought. Reload the page and the order does not shuffle to give someone a turn at the top. Determinism here is a promise that the tie-break is not a second auction wearing a mask.<\/p>\n
Notice what the ordering does not include. It does not include how recently a firm paid us extra. It does not include a hidden boost we sell off the menu. It does not include whether a firm complained loudly enough to get bumped. Tier, then validated rating, then a fixed tie-break. Three inputs, disclosed, in that order.<\/p>\n
We think disclosure of the ordering is as important as the ordering itself. A ranked list with secret rules is a slot machine. A ranked list with published rules is a scoreboard. You may not like that paid tiers sort first, but you can see the rule, weigh it, and read the page with that rule in mind. That is a different relationship with a reader than pretending the sort is neutral while quietly selling the top three positions.<\/p>\n
We offer a lifetime Premium option, and lifetime creates a problem that most subscription products never have to face. What happens when a firm buys lifetime placement and then, years later, the named attorney is disbarred.<\/p>\n
Our answer: the lifetime Premium is suspendable on disbarment, and we do not refund it.<\/p>\n
Let us walk through why. Lifetime, to a customer, sounds like a permanent guarantee of prominence. But prominence on a directory whose entire value is trust cannot survive the collapse of the underlying credential. If a lawyer is disbarred and we keep boosting that lawyer to the top of the placement rows because they prepaid, we have converted lifetime Premium into a shield against consequences. That is precisely the thing a trust directory cannot allow, because the whole reason a reader trusts the placement is the belief that we would not elevate someone the state has stripped of their license.<\/p>\n
So when disbarment is recorded against a lawyer whose firm holds lifetime Premium, we suspend the placement benefit. The profile can remain, often with the disciplinary status shown, because accurate information about a disbarment is useful to the public. What goes away is the premium elevation, the boosting of that firm above others as though nothing happened.<\/p>\n
No refund. This part sounds harsh until you sit with it. A refund would mean we owe money back because a lawyer got themselves disbarred, which frames disbarment as our failure rather than theirs. We did not cause the misconduct. We sold placement in good standing, and standing ended through the customer’s own conduct and the state bar’s action. The suspension protects readers. Refunding it on top would make the reader-protecting act into a cost we absorb for the disbarred lawyer’s benefit, and we are not going to structure it that way.<\/p>\n
Disbarment is not hypothetical or rare. State bars discipline lawyers constantly, and some of the names are large. Michael Cohen, President Trump’s former attorney, was disbarred in New York following his felony guilty pleas, a matter of public record from 2019. Rudolph Giuliani had his New York law license revoked; the Appellate Division, First Department disbarred him in 2024 over false statements about the 2020 election, and the District of Columbia moved against his license in the same period. John Eastman faced a California State Bar recommendation of disbarment in 2024 arising from the same post-election conduct. Jenna Ellis was censured in Colorado in 2023 after admitting to misrepresentations. These are people who, at the height of their public profile, could have paid for any placement money could buy. A directory that let prepayment override discipline would have kept elevating them past the moment their standing failed. We would not, and the lifetime terms say so before anyone buys.<\/p>\n
The general point sits under the specific one. A lifetime commercial promise cannot outrank a live credential. When the two conflict, the credential wins, and the money already paid does not buy its way back to the top.<\/p>\n
Now the part we care about most, because it explains the architecture rather than just describing it.<\/p>\n
Pay-for-trust is an incentive trap. The moment a directory lets payment produce a trust signal, the directory’s revenue starts pulling against its accuracy. Every fail becomes a lost sale. Every rejected firm becomes a customer who takes their money to a competitor with looser standards. If the person who approves badges also carries a revenue target, the target wins often enough to rot the signal, and once the signal is rotten, the badge means nothing, and once the badge means nothing, there is no reason to pay for it, and the whole thing hollows out from the inside.<\/p>\n
We have watched this play out across adjacent industries and it is not subtle.<\/p>\n
The credit rating agencies are the textbook case. Issuers paid the agencies to rate their securities, and in the run-up to 2008 the agencies handed AAA ratings to mortgage-backed instruments that were nothing of the sort. The Financial Crisis Inquiry Commission, in its 2011 report, called the failures of the rating agencies essential cogs in the wheel of financial destruction. The Senate Permanent Subcommittee on Investigations documented internal messages where analysts joked that a deal could be structured by cows and they would rate it. The rating was for sale, and the market eventually learned the rating meant nothing.<\/p>\n
Wells Fargo built a different flavor of the same trap. When you pay staff for account openings and punish them for shortfalls, staff open accounts. From 2016 onward the bank admitted that employees created millions of accounts customers never authorized, and the Consumer Financial Protection Bureau’s actions, including a large 2022 settlement, traced the fraud straight back to the incentive design. Reward the number and the number arrives, real or not.<\/p>\n
The auto industry gave us the mechanism in silicon. Volkswagen’s diesel software detected when it was being tested and cleaned up its emissions only for the exam, a fraud that surfaced in 2015 and led to guilty pleas and billions in penalties. When passing the check is the goal, someone eventually games the check instead of meeting the standard the check was meant to measure.<\/p>\n
Each of these is the same disease with different symptoms. A signal that was supposed to mean something got wired to a payment or a target, and the wiring corrupted the signal. We read those stories as engineering warnings, not distant scandals, and we designed against them at the level of the schema.<\/p>\n
Three design choices come straight out of that reading.<\/p>\n
We separated the tables. Billing writes to billing. Verification writes to verification. The badge template queries verification and cannot see billing. There is no join in the display path that would let a payment light up a badge. This is not a policy that a stressed manager can quietly bend on a bad revenue month. It is structural. To make payment produce a badge, someone would have to rewrite the code and leave the change in version history where the next person can find it.<\/p>\n
We kept a human at the publish gate and let the gate say no to paying customers. If publication were automatic on payment, the reviewer would be ornamental, and an ornamental reviewer is the first thing cut when volume climbs. By making the reviewer’s yes a required step that revenue cannot skip, we made the check load-bearing. The cost is speed. Paying firms sometimes wait, and they sometimes complain about waiting, and we accept both because the alternative is a queue that publishes whatever pays.<\/p>\n
We disclosed the ordering so we cannot secretly sell what we claim not to sell. A hidden boost is the easiest revenue in this business and the fastest way to corrupt a listing. By publishing tier, then validated rating, then deterministic tie-break, we removed our own ability to quietly insert a fourth, purchasable variable without visibly contradicting our stated rules. Disclosure disciplines the discloser. That is most of why we do it.<\/p>\n
The uncomfortable truth is that we make less money this way. A firm that fails verification is a firm we could have badged for cash. A page we hold in review is a dofollow link we could have published on payment. A disbarred lawyer’s lifetime placement is revenue we suspend and do not refund. Each of these is a decision to earn less in exchange for a signal that still means something a year from now. We think a badge that means something is the only asset we actually have, and the fastest way to destroy it is to sell it.<\/p>\n
We do not think we are morally superior to the directories that sell placement and call it trust. We think we made a different bet about what lasts.<\/p>\n
The bet is that readers eventually learn what a signal is worth. Not immediately. A pay-for-badge directory can run for years before the market catches on, collecting subscription revenue the whole time. But the catch-up arrives. The rating agencies enjoyed decades of authority before 2008 stripped it. The market found out what the AAA was worth, and it was worth roughly nothing on the instruments that mattered. We would rather be the boring directory whose badge still means the same thing in ten years than the profitable one whose badge quietly stopped meaning anything and whose readers have not finished noticing.<\/p>\n
So we keep the seams visible. Payment moves you up the rows and, once an editor publishes, gives you a dofollow link. Verification, and verification alone, lights the badge, from a stored row a human filled in against real documents. Ordering is tier, then validated rating, then a fixed tie-break, printed where you can read it. Lifetime Premium yields to a disbarment and does not pay itself back. Nothing scraped and nothing machine-written reaches the public without a person clearing it first.<\/p>\n
The Ohio firm that started this, for the record, sent their bar documents the following week. Their verification row filled in. Their badge rendered, from the verification table, exactly as it was built to. They had paid for placement on the day they upgraded. They earned the badge on the day the documents cleared. Two separate days, two separate systems, and that gap between them is not a flaw we are apologizing for. It is the product.<\/p>\n","protected":false},"excerpt":{"rendered":"
Paying and being verified are two separate things here. We explain what each tier buys, what it can never buy, and the incentive problems we built the whole system to resist.<\/p>\n","protected":false},"author":1,"featured_media":68,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[6,12],"class_list":["post-12","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-verification","tag-methodology","tag-transparency"],"yoast_head":"\n