Open banking's free front door just grew a price tag

What open banking is, and what quietly moved this month#

Every consumer fintech funnel has one step that has always been free. Connect your account. The customer taps a bank logo, signs in, and the data arrives before the coffee cools. That single tap is the quiet quarry of product design in consumer finance, and its price has been zero by default for so long that most teams have stopped seeing it as a price at all.

Open banking is the arrangement that lets a customer authorise a third party to reach the data held in their bank account. In the United States it has a statutory den: section 1033 of the Consumer Financial Protection Act, which the Consumer Financial Protection Bureau implements under the banner of Personal Financial Data Rights. The Bureau's page describes its October 2024 final rule as requiring data providers to make covered data available in an electronic form, subject to a number of requirements.

The statute underneath all of it is short and older than the argument. Codified at 12 U.S.C. 5533 and headed consumer rights to access information, it obliges a covered person to make account information available to the consumer on request, including costs, charges and usage data, and adds that the information shall be made available in an electronic form usable by consumers, per the United States Code. Read it closely and the silence is the story: the text sets no price and forbids none.

The 2024 rule that filled that silence was published on 18 November 2024 at 89 FR 90838 under docket CFPB-2023-0052, with an effective date of 17 January 2025 and RIN 3170-AA78, according to the Federal Register. Note the number. The reconsideration travels under a different RIN entirely, which is a filing detail worth carrying into any conversation where somebody insists the old rule is simply being edited.

What moved this month was procedure rather than principle. A proposed rule titled Personal Financial Data Rights Reconsideration, carrying RIN 3170-AB39, arrived for executive order review with a receipt date of 4 August 2026 in the record published by the Office of Information and Regulatory Affairs. Some trade coverage dates the submission two days later. The register field is the authority here, and the field reads 4 August.

No abstract is published on that review page. Nobody outside the building has read the text, review can run up to ninety days, and the wording stays in the burrow until Federal Register publication. The honest position on open banking fees today, then, is that the number does not exist in public and anyone quoting you one is guessing.

The scent was laid a full year ago#

This did not arrive from nowhere. An advance notice under the same RIN was published on 22 August 2025 in the Federal Register, docket CFPB-2025-0037, with comments closing on 21 October 2025. Its abstract states that the Bureau is considering the optimal approach to the assessment of fees to defray the costs incurred by a ‘covered person’ in responding to a customer driven request.

Read that line twice. A full year before the reconsideration reached review, the fee question was formally opened, and the proposal now waiting behind the door is the answer to it. The trail is a year long and it was walked in daylight, so anybody surprised this week was not tracking the ground. A patient prowl through the docket would have shown as much last autumn.

None of this lands on an empty field, either. The plumbing has its own industry body: the Financial Data Exchange convenes over 200 financial institutions, fintechs, data aggregators and industry leaders around shared standards for exactly this kind of access.

Why the fee question is really a funnel question#

Under the 2024 rule, data providers were broadly barred from charging for access. The reconsideration is widely expected to unwind that ban and permit reasonable fees to authorised third parties, most likely after some allowance of free requests, according to analysis published by Consumer Finance Monitor. That expectation belongs to the law firm, not to the Bureau, because the text is still sealed.

How firmly free was the old settlement? The Bureau's own announcement on 22 October 2024 said financial providers must make the information available without charging fees, and described the transfer to another provider at the consumer's request as free, in the release published by the Consumer Financial Protection Bureau. Free was not an accident of the drafting. Free was the sales pitch.

The banks have never been quiet about disliking that pitch. The Bank Policy Institute argues that banks should be able to charge third parties who seek access to that sensitive data, just as companies charge one another for products and services routinely in the marketplace, a case it set out in a one-pager in July 2025 and pressed in litigation filed on 22 October 2024 in the US District Court in Lexington, Kentucky. Whatever you make of the argument, it tells you which way the pressure has been pushing all along.

Take the expectation at face value and the consequence stops being abstract. A free step with a price is no longer a step, it is a line item. Onboarding acquires a variable cost that scales with curiosity rather than with conversion, and curiosity is the cheapest, most abundant thing on the internet.

Consider who actually taps the bank button today. Tyre kickers tap it. Comparison shoppers tap it. People who wanted a peek at the dashboard before deciding tap it. Under free access all of that browsing was a rounding error. Under priced access, every idle tap becomes a small invoice with no revenue standing behind it.

This is the fintech version of a very familiar problem, and folkfox has written about its cousins more than once, from the name-check screen that stops being a differentiator to the appointed representative acquisition chain. Regulation rarely changes what you sell. It changes where the friction sits, and friction has always been a marketing cost politely pretending to be an engineering one.

There is a second-order effect worth naming. If access carries a cost, the aggregators and the banks both gain a reason to police volume, and the noisiest, least qualified traffic is the first thing anybody polices. Sloppy funnels will feel it before careful ones do, which is a rare case of good practice paying a direct dividend.

So price the shape rather than the number. Take last month's connection attempts, split them by the screen that produced them, and run three scenarios: free forever, a modest allowance with a charge beyond it, and a charge on every attempt. You will not get the fee right, and that is fine. The point is finding which screens turn hostile the instant an attempt costs anything.

What US search demand says about the data access cluster#

Before anybody argues about price, it is worth looking at what people are actually asking. Using DataForSEO on 10 August 2026, folkfox measured US search demand across this cluster as primary research, and the pattern is tidier than the politics.

The head term, open banking, returns 18,100 monthly searches at difficulty 34, with a twelve-month trend of -11.3%. The builder-facing variant with api on the end is far smaller at 590 searches and difficulty 17, yet it is trending +111.6%. The rules-facing variants move the same way: the rules term sits at 110 searches, up 76.9%, and the regulation term at 50 searches, up 60.0%. Section 1033 itself carries 590 searches at difficulty 3, down 53.5%.

The shape is the story. General curiosity about the concept is cooling while the practical questions are heating up hard. That is a demand curve migrating from what is this toward what will this cost me, and it is precisely the migration a pricing change produces before the pricing actually lands.

Where that leaves your content plan#

A page that explains open banking in general terms is now chasing a shrinking crowd at the highest difficulty in the set. A page that answers what the connection will cost a product team, with numbers where numbers exist and candour where they do not, stands directly in the path the demand is drifting toward. That is a content marketing decision every bit as much as an SEO and GEO one.

One caution on the small terms: low difficulty and low volume together mean you win the query cheaply and win few clicks. Treat them as proof of positioning, not as a pipeline.

Sequence is the lever, and it is the one thing you fully control#

Here is the folkfox position, stated plainly. Nobody can tell you the fee. Everybody can tell you where in their funnel the connection gets requested, and remarkably few teams have ever questioned that placement, because until now it cost nothing to get it wrong.

Ask early and you pay to acquire people who never convert. The connection request lands before intent exists, so you buy bank data on browsers, abandoners and the merely curious. Ask late, after intent is proven, and you pay only for the people worth paying for. Sequence is the lever, not the fee.

Sequence carries a second benefit that survives the rule entirely. Every request you move later is a request the customer meets with more context, which lifts completion rates and lowers support volume, because people who understand why you want their bank data behave far better than people simply confronted with a login box. Cheaper and calmer at the same time is a rare combination, and it costs a fortnight of patient product work rather than a permanent budget line.

The counter-argument is real and deserves a hearing. An early connection can lift activation, because a connected account is a committed customer, and friction removed early sometimes pays for itself. Fine. Measure it. If that lift survives a per-request cost, keep it and sleep well. If it does not, you have discovered that your activation rate was subsidised by a policy rather than designed by you.

The discipline is the same one that separates a decent PPC account from a wasteful one, and it runs through good paid social too: pay for proven intent, never for mere movement. Open banking simply extends that rule from the ad auction into your own product, which is an uncomfortable place to find a media principle.

There is a brand argument sitting underneath the mechanics as well. Asking for bank access before you have said anything worth trusting is a positioning failure long before it is a cost failure, which makes this partly a brand strategy question about what you have earned the right to ask for, and when. The same trust arithmetic runs through AI financial advice and through stablecoin trust charters.

Building an open banking funnel that survives either outcome#

The fee could land at zero. A generous free allowance per customer per month would leave most consumer funnels essentially untouched, and a sensible reconsideration might do exactly that. Build for that outcome too, because a funnel rebuilt in a panic is a funnel rebuilt twice, and the second rebuild always costs more than the first.

Start with instrumentation, because feelings are expensive. Count connection requests, not merely successful connections, and split them by the screen that triggered them. Most analytics setups cannot answer that question today, which is why most teams will meet open banking pricing with anecdote rather than arithmetic.

It helps to look at how other regimes settled the same question. In the UK the Financial Conduct Authority states that providers of payment accounts accessible online must allow third-party providers to access user accounts with the user's explicit consent, under the Payment Services Regulations 2017 that first brought account information services and payment initiation services under regulation. In the EU, the payment services directive came into force on 12 January 2018 and is now being revised through a wider payments package, according to the European Commission. Two mature markets, both built on mandated access, neither collapsing.

Mandated access at scale is not a thought experiment either. UK account providers recorded 2,805,500,035 successful API calls in June 2026, a 99.50% success rate at an unweighted average availability of 99.35%, on data submitted by account providers themselves and published by Open Banking Limited. Hang any per-call price on a number shaped like that and the arithmetic stops being a rounding error, which is precisely why the American argument is being fought so hard.

The standards layer is not going anywhere while the rule is argued over, which is a small mercy for planning. Over 200 financial institutions, fintechs, data aggregators and industry leaders sit inside the Financial Data Exchange, so the rails you integrate against this quarter will still be the rails next year, whatever price rides on them.

Then build the fallback path properly. Manual entry, statement upload, or a soft-verified route that never touches the bank at all should exist and should convert, even if it converts worse. A single dependency with a price you do not set is a snare, and the moment to test the alternative is while the alternative is still optional.

None of that work needs the final rule. All of it is cheaper this month than it will be in the week the Federal Register publishes, when every product team in the sector starts the same job on the same deadline with the same scarce engineers.

Watch the register rather than the rumour mill, too. Review at the Office of Information and Regulatory Affairs can run up to ninety days, and the text becomes public at Federal Register publication, not before. Set a calendar reminder, not a Slack channel of speculation, and remember that the Consumer Financial Protection Bureau page carries no current fee position at all.

The wider pattern outlives this particular rule. Consumer finance keeps rediscovering that free infrastructure was a policy choice rather than a law of nature, and the brands that fare best are the ones that treated the free thing as a cost all along. Audience data taught the same lesson when first-party segments began quietly expiring, and measurement is teaching it again in today's piece on AI visibility tracking.

Regulated promotion carries the same structural risk, which is why the promotion risk piece sits beside this one on the same shelf. Read the pair together and the moral is consistent: the cheapest step in your funnel is the one most likely to be repriced by somebody who has never seen your funnel.

So take the vulpine view of it. Sit still at the hedgerow, watch the gate, and work out which way the wind is walking before that gate acquires a toll. Open banking has been free for long enough that free began to feel like a right, and the teams who prosper through the next eighteen months will be the ones who quietly costed the connection while everybody else was waiting for a headline number.

Frequently asked questions#