The fee is real. The verdict is not.
Something Awful’s forums have charged for registration for most of their existence. The current front page states the terms plainly: a one-time $10 price, roughly 150,000 Goons, and a rationale that reads like an operator wrote it after looking at a hosting invoice. “We charge money because it costs us money per month for bills alone, and since we don’t believe in shady internet advertising, we try to make the money back through forum registrations.” That is the entire public financial case, and it is worth taking seriously precisely because it is so thin.
What the record does not contain is any quantification. No revenue figure, no cost breakdown, no registration volume over time, no statement that the fee kept the site solvent. The popular framing — that the $10 gate “kept the forums afloat” — is a community belief, not an accounting result. For operators of 10k+ user boards, the useful question is not whether SA’s fee saved SA. It is what a one-time paid gate does structurally, what it cannot do, and how to decide whether it belongs in your own stack.
What the page actually says
The retrieved SA forums page gives us a small set of hard facts. Registration is a one-time $10 charge. The site says the money goes toward monthly bills and that it avoids advertising it considers shady. The page displays aggregate counters: 214,296 users total, 3,841,521 threads, 280,195,778 posts. The footer identifies the software as vBulletin Version 2.2.9 under a custom build string.
Wikipedia adds context that the page itself does not: that the site began charging $10 for forum access in 2001, and that it lists the registration fee as US$9.95 for posting privileges and full access, with additional features priced from $4.95 to $29.95. The $10 and $9.95 figures are both sourced and not reconciled here; treat the discrepancy as a reminder that even a single number on a checkout page can drift from the number in a reference entry.
Wikipedia also states that the site draws continuous income from new account fees, upgrades such as custom avatars and archive/search access, and merchandise. That is a description of revenue lines, not a measurement of them. It tells you the fee is one component of a mixed model, not the load-bearing wall.
A one-time fee is a stock, not a flow
This is the part operators get wrong most often. A one-time registration fee produces revenue at the moment of registration. It is a stock: a quantity that accumulates from discrete events. Hosting, moderation, and software maintenance are flows: costs that recur whether or not anyone registers this month.
If your board adds 200 paying registrants a month at $10, that is $2,000 in gross registration revenue for the month. If your recurring costs are $2,400, the gate does not cover them; it narrows the gap. If registrations fall to 40 a month because the community has matured and the backlog of interested newcomers is exhausted, the same gate produces $400. The cost line does not care that the fee was once sufficient. A mature forum’s registration curve tends to flatten, and a stock-based revenue model flattens with it.
This is not an argument against paid registration. It is an argument against modeling it as a subscription. If you adopt a one-time gate, budget it as a one-off contribution to fixed costs and pair it with a recurring mechanism — donations, a Patreon-style supporter tier, or advertising you can live with — unless you have evidence that registration volume will stay high indefinitely. SA’s own page lists Patreon among its purchase options, which is consistent with a mixed model rather than a fee-only one.
The moderation value is pricing re-entry, not stopping abuse
The common operator intuition is that a paid gate filters spammers and trolls. The SA record supports a narrower claim: the fee raises the cost of coming back. It does not eliminate return.
A 2025 thread in SA’s archives asks how responsible the $10 fee was in keeping the forums afloat. One poster recalls the early moderation regime: “there was no probation back in the day. So if you broke a rule, you got banned. It kept the forums really pristine. And if people wanted to be troublemakers they would just pay a price and keep going.” That is a description of a porous gate. The fee changed the economics of persistence; it did not close the door.
The same thread contains a poster claiming responsibility for the fee’s introduction, and other users treating the claim as forum lore. That is culture, not causation. The origin story is entertaining and it is not evidence. Operators should keep the governance lore and the financial ledger in separate drawers.
A 2014 thread asking users why they paid $10 to post on an internet forum is more useful as a catalog of motivations than as a cost-benefit study. Answers range from joining an EVE Online guild to supporting the site to one poster’s blunt framing: “I didn’t pay $10 to post on an internet forum. I paid $10 so those who don’t pay cannot.” Another poster tallied roughly $60 over the years across re-registrations, alt accounts, name changes, and avatars. That last detail matters for operators: the gate is not the only revenue surface. Re-entry, identity changes, and cosmetic upgrades are where a paid model actually accumulates.
The honest comparison is not free versus $10
For a 10k+ user board, the real choice is among four models, each with a distinct failure mode.
Advertising. Scales with traffic, not goodwill. It requires ad operations, privacy disclosures, and tolerance for the kind of inventory SA’s page explicitly rejects. It can work; it changes what your site is.
Donations and supporter tiers. Recurring if the community is loyal, volatile if it is not. It rewards existing members and does little to filter new ones. It also creates a visible class of supporters, which is a governance question as much as a financial one.
A one-time paid gate. Filters drive-by registration, funds a fixed cost at the moment of signup, and produces a stock that decays as the newcomer pool shrinks. It also creates a support burden: payment failures, chargebacks, gift accounts, and the inevitable requests for exceptions.
A hybrid. A modest gate plus recurring support plus optional upgrades. This is closest to what SA’s page describes, and it is the model most likely to survive a flattening registration curve.
The choice is not purely economic. SA’s stated aversion to “shady internet advertising” is a values decision with a price tag attached. Operators who share that aversion should price it explicitly rather than pretending the gate is free of tradeoffs.
A labeled hypothetical, not a finding
Consider a self-hosted board with 10,000 registered users, modest growth, and recurring monthly costs that the operator has actually measured. Under a $10 one-time gate, revenue equals registrations times ten. Under a donation tier, revenue equals the number of supporters times the tier price, which is a flow. Under advertising, revenue tracks page views, which is also a flow but depends on inventory the operator may not want.
None of these numbers are findings. The point of the exercise is that the gate’s revenue is the only line in the comparison that is a stock, and stocks require a registration-volume assumption that mature communities rarely satisfy. Run the comparison with your own measured costs and your own observed registration rate. If you cannot state the registration rate, you cannot budget the gate.
What to measure before you adopt a gate
Before charging for registration, collect four numbers. First, your actual recurring monthly cost, separated into hosting, software, and moderation labor. Second, your registration rate over the last twelve months, so you can see whether it is flat, rising, or decaying. Third, your abuse rate — spam accounts, ban evasions, and re-registrations — so you can tell whether the gate is solving a problem you have. Fourth, your support capacity for payment issues, because a checkout page is a support queue.
If the registration rate is healthy and the abuse rate is high, a gate may earn its keep as a filter. If the registration rate is flat and the cost line is the problem, a gate is a one-time patch on a recurring wound. If the abuse rate is low, the gate is a values statement more than a control.
What the SA record proves is narrower than the folklore: a paid gate can coexist with a large, long-lived forum, and its operators describe it as a way to cover bills without advertising they distrust. What the record does not prove is that the gate kept the site alive, that it stopped abuse, or that the same model will work at your scale. The fee is a design pattern. The verdict is yours to measure.
FAQ
Is the Something Awful registration fee $10 or $9.95?
Both figures appear in retrieved sources. The forums page states a one-time price of $10; Wikipedia lists US$9.95 for posting privileges and full access. The discrepancy is not reconciled by either source.
Did the fee eliminate re-registration and ban evasion?
No retrieved source supports that. SA forum threads describe re-registrations, paid unbannings, and users tallying multiple account purchases over time. The fee raises the cost of re-entry; it does not close it.
Can a one-time fee replace recurring revenue?
Not reliably. A one-time fee is a stock tied to registration volume, while hosting and moderation are recurring flows. Operators should budget it as a one-off and pair it with a recurring mechanism unless they have evidence of sustained registration volume.
Does the SA page say the fee kept the forums afloat?
No. The page states that the site charges because it costs money per month and that it avoids advertising it considers shady. No retrieved source quantifies the fee’s contribution to the site’s survival.
