Seventeen of 20 B2B fintech homepages we opened on 28 October 2026 named their product category in the hero, but only six said who the product is for. Fintech heroes say "payments", "spend" or "core banking" readily. They rarely say "for finance teams", "for merchants" or "for banks", even when the whole business is built around one of those buyers.
Below are the nine tests, every measurement, and what the top scorers do that others could copy. The tests are the ones we used for legal tech and B2B SaaS homepages earlier this month, so the samples compare directly.
The rubric: nine pass or fail tests
Each test passed scores one point, so each dimension scores 0 to 3 and the total is out of 9.
Clarity, three tests. C: the hero names a product category or core job in words buyers already use. A: it names who the product is for, as a role, team or company type. S: it gives one concrete specific beyond the category, such as named functions or a number. The basis is Nielsen Norman Group's first homepage guideline, which asks for a tagline that "summarizes what the site or company does", and applies it to well-known companies too. The same group's dwell-time analysis of 205,873 pages found the first 10 seconds decide whether a visitor stays.
Speed, three tests. Time to first byte of 0.8 seconds or less, web.dev's "good" line. HTML transfer of 155 KB or less, and of 78 KB or less: the 90th and 75th percentiles of mobile home page HTML in the Web Almanac's 2025 page weight chapter.
Demo path, three tests. A demo or sales form one click or less from the first screen; five fields or fewer on its first step; and a stated response time or call length.
Clarity method. The hero is the headline, the line under it and any label directly above, read at 1,024 by 768 pixels without scrolling. In fintech, "payments", "business banking", "spend" and "accounts payable" name a job buyers shop for, so they pass. "Financial infrastructure", "financial products", "fintech" and "agentic banking" do not, because a buyer cannot tell from them what the product does. Cookie banners and chat widgets were ignored.
Speed method. Five curl requests per homepage between 11:09 and 11:10 UTC from one connection in Serbia, taking the median time to first byte and the median compressed size of the HTML document. Browsers usually receive brotli, which is smaller, so our sizes err high.
Demo path method. Six heroes lead with self-serve sign-up: Stripe, GoCardless, Ramp, Brex, BILL and Mercury. To compare like with like, we scored the sales or demo form on every site, followed from the most prominent sales link in the first screen, and counted visible fields on the first step in a rendered browser at 1,280 pixels. A phone number with a country picker counts as one field; consent boxes count as none. We submitted nothing.
The scorecard: clarity and speed
Letters show which clarity tests a hero passed. Platform comes from our CMS check of 93 fintech vendors, published as Fintech Website Platform and CMS Statistics 2026.
| Site | Product | Platform | Clarity | Passed | TTFB (s) | HTML (KB) | Speed |
|---|---|---|---|---|---|---|---|
| Stripe | Payments | Next.js | 2 | CS | 0.33 | 190 | 1 |
| Adyen | Payments | Nuxt | 1 | C | 0.28 | 330 | 1 |
| Checkout.com | Payments | Webflow | 1 | C | 0.11 | 69 | 3 |
| GoCardless | Payments | Framer | 2 | CS | 0.18 | 271 | 1 |
| Ramp | Spend management | Next.js | 2 | CS | 0.80 | 129 | 1 |
| Brex | Spend management | Next.js | 2 | CS | 0.20 | 75 | 3 |
| BILL | Payables and receivables | Webflow | 2 | CS | 0.15 | 69 | 3 |
| Tipalti | Payables and payouts | WordPress | 3 | CAS | 0.42 | 46 | 3 |
| Pleo | Spend management | Astro | 2 | CS | 0.23 | 30 | 3 |
| Marqeta | Card issuing | Next.js | 1 | C | 0.16 | 109 | 2 |
| Unit | Embedded banking | Webflow | 1 | S | 0.17 | 259 | 1 |
| Thought Machine | Core banking | Webflow | 2 | CA | 0.39 | 49 | 3 |
| Mercury | Business banking | Next.js | 2 | CS | 0.22 | 164 | 1 |
| Alloy | Identity and fraud | Craft CMS | 2 | CS | 0.30 | 109 | 2 |
| Persona | Identity verification | Astro | 1 | C | n/a | n/a | n/a |
| Sardine | Fraud and compliance | Next.js | 3 | CAS | 0.21 | 124 | 2 |
| Plaid | Open banking | Next.js | 0 | none | 0.57 | 42 | 3 |
| TrueLayer | Open banking | Gatsby | 2 | CA | 0.12 | 54 | 3 |
| nCino | Lending | Next.js | 2 | AS | n/a | n/a | n/a |
| FloQast | Close management | Webflow | 2 | CA | 0.07 | 50 | 3 |
Source: Octo Agency, rendered browser read and five curl runs per homepage, 28 October 2026. Persona answered curl with a bot challenge (HTTP 403) and nCino with HTTP 429, so their speed is n/a. Ramp's median was 0.8002 seconds.
The scorecard: demo path and totals
| Site | Form | Clicks | Fields | Timing stated | Demo path | Total |
|---|---|---|---|---|---|---|
| Stripe | Sales, 3 steps | 1 | 2 | No | 2 | 5 |
| Adyen | Sales, 2 steps | 1 | 2 | No | 2 | 4 |
| Checkout.com | Sales, 2 steps | 1 | 7 | No | 1 | 5 |
| GoCardless | Sales | 1 | 7 | No | 1 | 4 |
| Ramp | Demo | 1 | 1 | Yes, 30 min | 3 | 6 |
| Brex | Demo | 1 | 5 | No | 2 | 7 |
| BILL | Demo | 2 | 4 | No | 1 | 6 |
| Tipalti | Demo pop-up | 1 | 5 | No | 2 | 8 |
| Pleo | Demo | 1 | 6 | No | 1 | 6 |
| Marqeta | Sales | 1 | 7 | No | 1 | 4 |
| Unit | Sales, 2 steps | 1 | 4 | No | 2 | 4 |
| Thought Machine | Demo | 1 | 10 | No | 1 | 6 |
| Mercury | Sales | 2 or more | 9 | No | 0 | 3 |
| Alloy | Demo | 1 | 7 | No | 1 | 5 |
| Persona | Demo | 1 | 3 | No | 2 | 3 of 6 |
| Sardine | Demo | 1 | 8 | No | 1 | 6 |
| Plaid | Sales, on homepage | 1 | 6 | No | 1 | 4 |
| TrueLayer | Sales | 1 | 5 | No | 2 | 7 |
| nCino | Discovery call | 1 | 9 | No | 1 | 3 of 6 |
| FloQast | Demo pop-up | 1 | 7 | No | 1 | 6 |
Source: Octo Agency, rendered browser read of each form, 28 October 2026; no form was submitted. Fields counts the first step only
Clarity: fintech names the product and skips the buyer
The category test was the easy one. Payments companies say payments, spend companies list cards, expenses and bills, and core banking vendors say core banking. Only Unit, Plaid and nCino failed it. Unit's "Build financial products on real infrastructure" and Plaid's "revolutionary financial products" describe what the customer will make, not what the vendor sells. nCino's "Agentic Banking Lives Here." names a technology trend.
The buyer test is where fintech differs from the other samples. Six heroes named one. Tipalti's "AI built for finance" and FloQast's "AI built by accountants for accountants" name a team. Thought Machine names "tier 1 global institutions" and "hyper-growth fintechs"; Sardine names "banks, merchants, and fintechs"; TrueLayer names merchants twice; nCino names "your bankers". The other 14 say "your business", "the world's leading businesses" or nothing. That is a choice with a cost: Stripe, Adyen and Checkout.com sell to very different merchants, and a visitor has to open the navigation to find out whether the site means them.
Numbers are fintech's favourite specific. Stripe's label reads "Global GDP running on Stripe" with a live percentage; Ramp's shows its share of US corporate payments; Brex says "120+ countries"; Mercury says "Apply online in 10 minutes". Twelve heroes passed the specific test, mostly on a number or a list of functions.
Tipalti and Sardine pass all three. Tipalti's line under the headline names accounts payable, mass payments, procurement, expenses and treasury. Sardine's names fraud prevention, AML compliance and real-time transaction monitoring, then its buyers. Sardine's headline animates letter by letter, so for the first second the screen reads "The agentic risk platform for", which is worth checking on any animated hero.
Regulation shows up in the first screen. Mercury puts a disclosure in its hero: it is a fintech company, not an FDIC-insured bank, and its partner banks are named. We did not score it, but it is the clearest example in the sample of compliance copy sharing space with the headline.
Speed: fast servers, heavy documents
Time to first byte separated almost nobody. Seventeen of 18 measured homepages answered within 0.8 seconds, with a median of 0.22 seconds. Ramp's five runs ran from 0.48 to 1.29 seconds and its median landed at 0.8002, just over the line. web.dev notes that TTFB is not a Core Web Vitals metric, so this sits behind the visitor's experience rather than measuring it.
Document weight is where fintech differs from legal tech. The median homepage sent 92 KB of HTML, against the Web Almanac's 33 KB median and 78 KB 75th percentile for mobile home pages. Five of 18 went over 155 KB, and the causes are specific.
Adyen embeds its data. Of the 1,178 KB uncompressed document, 969 KB is one Nuxt data script, the content the page already shows, serialised a second time.
Stripe, GoCardless and Mercury inline graphics. Inline SVG makes up 39% of Stripe's uncompressed HTML, 35% of GoCardless's and 22% of Mercury's. Stripe and Mercury also carry about a third of their document as Next.js page data.
Unit is not compressed. It sent 259 KB with no compression, even when our request accepted gzip and brotli. Turning on compression at the host would likely cut that by most of its size.
Platform was not the story. The five Webflow sites answered in 0.39 seconds or less and the eight Next.js sites within 0.57, apart from Ramp. We build marketing sites in Webflow, so weigh that with our bias in mind: a cached page on any modern host answers fast.
Demo path: short forms are rare, timing is rarer
Eighteen of 20 sites put a sales or demo form one click from the first screen. BILL and Mercury were the exceptions. At 1,024 pixels BILL shows only its sign-up button; Mercury shows "Open account" and an interactive demo, with the sales form further away.
Six fields is the fintech median too. First steps ran from one field to ten, and nine of 20 asked five or fewer. Ramp asks for a work email only, then moves to scheduling. Our fintech demo request statistics read 25 fintech forms eight days earlier and found a median of seven, on a partly overlapping sample; it covers phone and country questions in detail.
Money questions come early. GoCardless and TrueLayer ask for annual revenue on the first step, and Mercury asks for expected balance size. That is qualification, not friction for its own sake: a payments provider prices on volume. Our fintech landing page statistics discuss form length against lead quality.
The best published field-count data is thin and comes from a vendor. Factors.ai, which sells analytics, reports that demo forms with four to six compulsory fields had a 51% success rate against 30.60% for seven or more, from over 50,000 web sessions. Eight of our 20 forms asked seven or more.
Almost nobody says what happens next. Ramp promises a 30-minute tour. Tipalti tells visitors they will pick a time from a calendar after the form, which is useful but gives no call length. TrueLayer says an open banking expert "will reach out". Chili Piper, which sells scheduling software, reports from nearly 4 million form submissions that letting buyers book immediately after the form lifts booking from 30% to 66.7% on average.
How fintech compares with legal tech
The legal tech sample was read on 21 October 2026 with the same tests and the same five-run speed method.
| Measure | Fintech, 20 sites | Legal tech, 20 sites |
|---|---|---|
| Median total, of 9 | 5.5 | 5 |
| Hero names a category | 17 of 20 | 9 of 20 |
| Hero names the buyer | 6 of 20 | 8 of 20 |
| Clarity score of zero | 1 of 20 | 6 of 20 |
| Median time to first byte | 0.22 s | 0.10 s |
| HTML over 155 KB | 5 of 18 | 1 of 19 |
| Median first-step fields | 6 | 6 |
| Call length or response time stated | 1 of 20 | 3 of 19 |
Source: Octo Agency scorecards, 21 and 28 October 2026
Fintech heroes say what they sell far more often, carry heavier documents and say less about the call. The buyer gap is the cheapest of these to close.
What this means for a fintech website
Run the tests on your homepage and two competitors; each site takes about 30 minutes.
1. Add the buyer to the line under the headline. If your product is for finance teams, merchants or banks, say so in four words. It costs nothing and only six of 20 did it.
2. Weigh the HTML. Anything over 155 KB usually hides serialised page data, inline SVG or a missing compression setting. All three are fixable without a redesign.
3. Walk the sales path at 1,024 pixels. If your hero leads with sign-up, keep a visible sales link for the buyers who need a contract. Then count first-step fields, keep the revenue question if it routes the lead, and add one sentence on call length. Our fintech website conversion rate statistics give the benchmarks to measure the change against.
We would run these nine tests on staging before launching any fintech marketing site we build in Webflow, as we would for any B2B site. Our marketing website service explains how we scope that work.
