We Audited Our Own 70-Tool Pricing Database: 7 of 30 Prices Were Wrong

Every DIY-alternative guide on Memvers recommends specific tools — and every one of those tool cards, on every page, reads its name, price, and free-tier status from a single file: tool-registry.ts, 70 tools deep. A bridge function called resolveRecommendedTool() is what actually pulls the canonical data out of that file at render time instead of letting each guide keep its own hand-typed copy of a price. That bridge itself is a fairly recent fix — before it existed, a DIY entry could reference a tool by name and just retype whatever price felt current at the time it was written, with nothing forcing those numbers to agree with each other, let alone with reality. Wiring pages to read from the registry directly caught real, dollar-figure drift on three tools the first time it ran: CapCut, Canva Pro, and ElevenLabs.

That fix solved a consistency problem — every page now agrees with the registry. It didn't solve an accuracy problem: the registry itself can still go stale, the same way any pricing page does, because software vendors change prices, rename tiers, and quietly shrink free plans all the time. Nobody had gone back and asked whether the registry's 70 entries are still correct, as opposed to merely consistent. So we did — not as a survey of the whole file, but as a real, hands-on spot-check of a meaningful sample, with every finding checked against a live source before we touched anything.

tool-registry.ts is the single source of truth behind every DIY guide's tool recommendations on this site — a typed record per tool with a name, price string, free-option flag, category, and affiliate metadata. We didn't audit all 70 entries; we picked a sample of 30, deliberately weighted toward subscription software (design, AI, video, audio, ecommerce, automation tools billed monthly or annually) over one-time purchases and free/open-source tools, on the premise that a recurring-billing product can reprice at any point, while something like Blender or a $295 one-time DaVinci Resolve Studio license mostly can't. We included a small control group of exactly that kind — DaVinci Resolve Studio and Unreal Engine 5 — specifically to test whether the "subscription drifts more" assumption actually holds in our own data, not just assert it.

For each of the 30, we searched for the vendor's own current pricing page or a recently-dated third-party tracker, compared it line-by-line against the registry's exact price string, and classified the result as either confirmed accurate or drifted — and if drifted, we edited tool-registry.ts itself with the corrected figure, not just a note in this post. Every corrected entry in the file now also carries an inline comment recording what changed and what we verified it against, the same way we'd want a future audit (by us or anyone else) to be able to check our work.

Before this pass existed, the same three tools — CapCut, Canva Pro, and ElevenLabs — were the concrete catch that justified building the registry bridge in the first place: DIY guides were quoting stale, hand-typed prices for all three until they were wired to read from tool-registry.ts directly. Since that's the exact case this post's premise rests on, we re-checked all three against current pricing rather than assuming a past fix stays fixed forever.

These are the entries where we changed tool-registry.ts itself, not just flagged an issue. Every row below is a real edit, live in the source file that every DIY guide's tool cards read from.

Read the chart for what it actually says: more than half of the real drift we found (4 of 7) wasn't a number changing at all — it was a vendor adding a new tier somewhere in their lineup that made an otherwise-correct two-tier price string incomplete. Adobe Creative Cloud, CapCut, Suno AI, and Udio all fall into this bucket. The remaining three are more traditional: Mailchimp shrank what "free" means without touching the paid price, Helium 10 killed its cheapest paid plan outright and replaced it with something roughly triple the cost, and Epidemic Sound renamed its entry tier at the same time it raised the no-commitment price.

We're not going to only show you the mistakes. Here's the full sample, including the 23 tools where the registry's existing price string checked out and we made no edit at all — accuracy is the more common outcome in this data, not the exception.