
Web3 website design: how to build a crypto website people trust
Imran KhanNovember 18, 202613 min read

table of contents
Most Web3 websites get one chance to say three things: what this is, whether it's safe, and where the real app lives. Visitors arrive wary, often from a link in a chat, and the next click may open a wallet.
This guide is about designing that website. We cover which site you're really designing, the pages it needs, the homepage, where the wallet prompt belongs, how to show proof, how to defend against copycats, what your words are allowed to claim, speed and accessibility, what it costs, and how we'd run the project.
One thing up front. We're a design studio, and we haven't built a protocol's marketing site. Our Web3 work is Reputable, where we designed both the app and the website that brings people to it. We use it for what it shows, and rely on linked sources for the rest. Every outside fact was read on 18 November 2026. Nothing here is legal or financial advice.
The short version
- Decide which site you're designing. A marketing site, an app and a docs site have different jobs. Most guides mix them.
- Say what it does in plain words in the first screen, with no "scalable, secure, compliant" filler.
- Keep the wallet prompt off the first screen unless the visitor asked for it. That's our advice, and the reasons are below.
- Give proof with dates and scope. An audit badge means little unless the page says what was audited.
- Build an official links page and put the "we will never ask you to" list on it.
- Treat third-party scripts as part of your security. The Ledger Connect Kit incident came through a code library, not a design flaw.
- Check your claims against the rules. In the UK, crypto promotions have had to be clear, fair and not misleading since October 2023.
- Hold the site to numbers: 2.5 seconds for the main content, 200 milliseconds to respond, 4.5 to 1 text contrast.
First, which website are you designing?
Search for this topic and you'll find guides that treat three different things as one. Clay writes about wallets, signing and recovery, which is app design, and its FAQ defines a Web3 website as a decentralized application. Merge writes about hero copy and tech stacks, which is marketing site design. Both are useful. They're not the same job.
| Marketing site | The app | |
|---|---|---|
| Who arrives | Someone who has heard of you and isn't sure | Someone ready to act |
| The job | Explain, prove, point to the real app | Let them do the thing safely |
| Wallet involved? | Rarely, and only when asked | Always |
| Where to read more | This guide | Our Web3 UX guide and crypto app design guide |
A third site, the docs, answers builders and users who hit a problem without a support ticket. It needs the same care and its own owner.
The split matters because the marketing site is the page copycats clone first. If it's also where visitors are asked to connect a wallet, a fake and the real one look the same to someone in a hurry.
What changes when the visitor might be holding money
A normal marketing site loses a sale if it fails. A Web3 site can cost a visitor their funds if they trust the wrong page. That shifts what the design is for.
Three things follow.
The visitor is checking you, not just reading you. Clay advises showing relevant stats "with dates and sources". Lazarev compares good Web3 design to an ATM: people don't want to think about what's behind the scenes, they want it to feel safe. We'd go one step further. Every claim on the page should be something a skeptic could check from the page itself.
Someone is almost always imitating you. In October 2026 Malwarebytes reported 70 sites impersonating legitimate crypto projects, copying the logo, menus and colours. Its "Vote now" button opened a wallet connection prompt. Its advice to visitors was to check any claimed vote or rewards distribution through the project's official channels before connecting a wallet. The site you design has to make that check easy.
Your words are regulated. More on that below.
The pages a Web3 website needs
The ranking guides list features. Here's the list of pages we'd start from, and what each one is for.
| Page | Its job |
|---|---|
| Home | What it is, who it's for, proof, the way into the app |
| How it works | The mechanism in plain words, with one worked example |
| Security | Audits with scope and date, bug bounty, what you do and don't hold |
| Fees and risks | What a user pays and what can go wrong, written for a first-time user |
| Token or economics, if you have one | What it does, supply, who holds it, stated as facts with dates |
| Docs | For builders, linked from the nav |
| Team and backers | Real names where you can, and why it matters that people can be found |
| Official links | The one list of real addresses and accounts |
| Blog or updates | Dated changes, so the site shows the project is alive |
| Legal | Terms, privacy, and the risk disclosures your lawyer requires |
Designmonks lists tokenomics and whitepaper presentation as core features. We agree it needs a home. Whatever you publish there, date it.
The homepage
Say what it is in one sentence a stranger understands
Ethereum.org opens with "The internet that belongs to you" and follows it with "Ethereum is the global network where you control your assets, your data, and your identity." You may not like the claim, but you can read it in four seconds. Compare that with the line Merge warns against: overused phrases such as "scalable, compliant, and secure".
Here's a test we'd run on any Web3 hero. Cover the logo and show it to someone outside crypto. Can they say what the product does and who it's for?
The picture above is a mockup from our Reputable project. The sentence on the poster names the product, says what it does and says who it's about. The case study says Reputable was built on one idea, "your work should speak for you", so the line didn't need decoration.
Give the proof next to the claim
If the hero says something checkable, put the check beside it. Look at how three live homepages handle it, as we saw them on 18 November 2026:
- Lido has a Security section with a list of auditors and a "See all audits" link.
- Aave says "Independently audited, onchain, and open to verify."
- Safe states "18+ independent audits" on its security page.
These are the projects' own statements, and we haven't checked the audits themselves. The design point is where they put them: on the page, where a visitor can see them without hunting.
Our advice is to go further than a badge. Say what was audited, by whom, when, and whether the code has changed since. Clay puts it in its editorial checklist: clarify what any audit really covered, and date any market number.
This is what the Reputable case study calls "show the working". Every number in the product is one click from its evidence: the commit heatmap, the highlighted commits, the protocols a developer shipped to. That's an app principle, and we'd hold a marketing site to the same standard: a figure on the homepage should lead somewhere.
A first-screen checklist
We can't show a protocol site we designed, so here's the layout we'd draw, in order from the top:
- A name and a one-sentence description a stranger can repeat.
- One main button that says where it goes, such as "Launch app", and one quieter link to the docs.
- One proof line with a date and a link to its source.
- A link to the security page in the main navigation.
- The official links page in the footer.
Everything else, such as the animation and the partner logos, comes after those five.
Pick one main action and name it honestly
Most Web3 homepages have a button that says "Launch app", "Connect wallet" or "Get started". We prefer the first, because it says where the click goes. "Connect wallet" on a marketing page promises something the visitor may not be ready for. The next section says why.
Where the wallet prompt belongs
Our advice: don't ask for a wallet on the marketing site unless the visitor has chosen an action that needs one. Three reasons.
- A fake can copy a prompt perfectly. The Malwarebytes sites worked because a believable button opened a believable window. If your real site teaches visitors to connect the moment they land, a clone can use that habit against them.
- It's an early demand for trust. Clay recommends building trust with small wins and starting with low-risk actions. It's lower risk to read a page than to connect to it.
- Wallet support is messier than one button. The EIP-6963 standard exists to discover multiple injected wallet providers in a web page, so a site isn't built around whichever wallet grabs the single browser slot. If you do offer a connect button, use a library that supports it and let the visitor choose.
When a signature is needed, Clay has the right instinct: before signing, show who is asking, for what purpose and for how long. That's an app screen, and our Web3 UX guide covers it.
Trust that survives a skeptic
A security page that says what you don't do
A security page is a list of facts: audits with dates and scope, the bug bounty if there is one, who holds keys, what can be paused or changed and by whom. Write the limits too. If the contracts are upgradeable, say so.
Add a short "we will never" list. The model is wallet support. MetaMask's help centre says never to share a Secret Recovery Phrase with anyone, even someone who says they're from MetaMask Support. Your own list will differ, but it should be this blunt, and it should be easy to find.
An official links page
One page, one URL, linked from the footer and the docs, listing the real website, app, docs, contract addresses where relevant, and the real social accounts. Put the same URL in every profile. When a visitor doubts a link someone sent them, this is where they check.
Our advice is to treat it as a design object, not a footer afterthought. Large type, no decoration, one list, a "last updated" date, and a note on what the page is for.
People
If your team is anonymous, say how a visitor can still judge the project: public code, audits, a track record on chain. If your team is public, show it. Both choices are fine as long as the page doesn't hide which one it is.
Protect the site itself
For a Web3 site we think this belongs in the design brief, because a hijacked page is a brand failure.
Third-party scripts. In December 2023, Ledger reported that its Connect Kit, a JavaScript library that adds a button for connecting a wallet to a site, had been compromised. A former employee had fallen for a phishing attack, which let an attacker upload a malicious file to the NPM package registry. The affected versions were 1.1.5, 1.1.6 and 1.1.7. Ledger said a fix was deployed within 40 minutes of its becoming aware, that the malicious file was live for around five hours, and that the exploit was effectively running for less than two hours. The lesson for a website team is that code you load can change after you launch.
Here's what we'd do about it. This is our advice, not Ledger's. Keep the number of third-party scripts on any page with a wallet button as low as you can. Pin the versions you use, and check what changed before you update them. Ask your developers whether they bundle libraries into the site's own build or load them live from someone else's server.
Domains. Our advice: register the obvious look-alikes of your name and the common typos, and point them at the real site.
A page that tells people about incidents. Our advice: if something goes wrong, the site is where people look first. Decide now where the notice goes.
Words: what the site is allowed to say
If the project sells or promotes a crypto asset to people in the UK, the website copy may count as a financial promotion. The FCA announced rules that, from 8 October 2023, require clear risk warnings and adverts that are clear, fair and not misleading, and a cooling-off period for first-time investors. The page was last updated on 6 February 2026. Other countries have their own rules.
We aren't lawyers, so this is about what to design around, not what's legal.
- Plan space for risk wording. A warning pushed into small grey text at the bottom isn't a design decision a regulator will like. Design it in from the start.
- No invented urgency. Pressure is a scammer's tool: the Malwarebytes write-up describes a fake page adding fake dates and a countdown. A real site that does the same looks like the fake.
- Date every figure. Total value, users, yield: if it moves, say when it was read.
- Don't promise returns. Describe what the product does, and let the numbers speak with their dates.
- Have counsel read the page. Before launch, not after.
For the brand side of this, see our Web3 branding guide.
Look, motion and mobile
The guides name dark backgrounds, 3D and motion as current Web3 trends. Merge names 3D, dark mode, minimalism and bold abstract compositions as current trends, and The Alien adds interactive elements and mobile. We'd keep two limits.
Contrast is a number. WCAG 2.2 asks for a contrast ratio of at least 4.5 to 1 for text, with some exceptions. Mid-grey type on near-black is an easy way to miss it on a dark site. Measure, don't judge by eye.
Motion should explain something. An animation that shows how a swap works earns its place. A background that makes the main content load later doesn't.
Type is one place to keep it plain. For Reputable we kept one type scale and a dense, calm layout on purpose, as the case study puts it, for people who read tables all day. A clear typeface set at readable sizes does more for trust than another effect.
Design desktop and phone in the same pass. Links get opened from chats and social posts, and the first screen has to work on a small one.
Speed
Google's web.dev gives three targets, measured at the 75th percentile of page loads across mobile and desktop: Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint of 200 milliseconds or less, and Cumulative Layout Shift of 0.1 or less.
Heavy 3D heroes, video backgrounds and a stack of wallet and analytics libraries are the things we'd check first when a page misses them. Merge recommends Next.js or Webflow for a fast and cost-efficient site. Our own site is built on Next.js, and we build client sites in Webflow and Framer, so we'd choose with you on the first call.
How to test it
Clay recommends testing across the main wallets, browsers and devices, and measuring confidence rather than clicks. For the marketing site, here's what we'd add:
- Show the first screen to five people outside crypto and ask what the product does.
- Ask a user to find the audit report and the official links. Time them.
- Open every page on a real phone on a slow connection.
- Run the real site and one of your look-alike domains side by side, and ask whether a visitor could tell which is which.
- Re-check dates and figures before every launch.
What it costs and how long it takes
Prices vary a lot, so here are other agencies' published ranges, labelled as theirs. We don't print our own prices in articles: the pricing page explains how we charge.
| Source | What it says |
|---|---|
| Designmonks | Marketing sites $5,000 to $15,000; Web3 product sites $15,000 to $40,000; full dApps $40,000 to $100,000+ |
| Avark | As little as 2 weeks for design only and 4 weeks for design and development, depending on the spec |
Those are their estimates, not ours, and a regulated or audited project can sit outside them. For Usarion, our pricing page explains how we work: a fixed price for a defined scope, or a monthly plan. Our SaaS website design service says four to eight weeks for most sites, with the homepage visible in the first week.
How we'd run the project
This is the process from our own service page, adapted to a Web3 brief. It's a plan, not a record of a Web3 marketing site we've shipped.
- Work out the pitch (two to three days). Who it's for, what they compare you with, and the one sentence that has to land. Add: what must be provable.
- Structure the pages. The page list above, each with one job. Agreed in writing before design.
- Design with real copy. Desktop and mobile together, with the risk wording and the official links page in the first drafts.
- Build and wire up the CMS. So your team can publish dated updates without a developer.
- Launch and measure. Redirects, analytics and forms tested, then fix the page that leaks. For a Web3 site, also the look-alike domains and the incident notice.
If a site's job is to explain a product and get people into it, our SaaS website design best practices covers the parts that aren't specific to Web3.
A worked example: Reputable
Reputable is a reputation platform for Web3 developers. We designed it from the first sketch to launch over three months: the web app, in dark and light, and the website that brings people to it.
Three things from that project carry over to any Web3 site.
- Every number leads somewhere. The case study says the team needed a product that made their score believable, so every figure sits one click from the work behind it.
- Claim, don't sign up. A profile exists before its owner arrives, and signing in with GitHub claims it. The site's job is to get that first click right.
- One design system for the team. The case study says we worked from the founders' ranking model to a design system ready for their developers, and drew every screen in dark and light.
According to the case study, in the first three months after launch developers claimed profiles faster than the founders planned, with 9.4k profiles claimed, a 3.2× figure for profiles seen per visit, and 22 screens in both themes. Those are the figures on that page. We haven't published a breakdown of how much of that came from the website.
Common mistakes
- The app and the marketing site handled as one project.
- No official links page, so visitors can't check anything.
- Third-party code from servers you don't control, on pages with wallet buttons.
- Countdown timers and returns language.
- Grey text on dark backgrounds that fails contrast.
- A 3D hero that makes the main content wait.
FAQ
The term is used two ways. Some guides mean a decentralized app that runs on blockchain, and others mean the marketing site of a Web3 project. This guide is about the second, the site that explains and proves the project and points to the real app.
Visitors may be about to use real money, and imitators are likely. So the site leans on proof, clear claims, an official links page and a careful approach to wallet prompts.
Home, how it works, security, fees and risks, docs, team, official links, updates and legal. A token page too, if there is one.
We'd wait until they choose an action that needs one. The reasons are in the wallet section above.
For a marketing site, no. Nothing in the guides or standards we read requires it. What matters for the design is whether a visitor can trust the page, and that comes from proof, official links and a secured build.
No. Use the same standards: 4.5 to 1 text contrast under WCAG 2.2, and the web.dev targets for loading, responsiveness and layout shift.
It depends on scope. Designmonks publishes $5,000 to $15,000 for a marketing site and $15,000 to $40,000 for a Web3 product site. Usarion's approach is on the pricing page.
Avark says as little as two weeks for design only and four for design and development. Our SaaS website service says four to eight weeks for most sites.
We can design and build the site, and we've designed a Web3 product and its website in Reputable. We haven't built a protocol's marketing site, so ask us for the exact work you need.
If you're building a Web3 product and want the website designed so the claims are checkable, the wallet prompt sits where it should, and the official links are easy to find, book a call. Our SaaS website design work covers the site from the page list to launch. The list of Web3 design agencies shows who else does this.
About the author
Imran Khan, Founder & Creative Head
15 years of industry experienceFormer Head of Design at Day TranslationsFormer Creative Lead at SupersideWork at Superside for brands including Meta, Red Bull and Ray-Ban

related articles
Book a call
Pick a time that suits you
Get in touch
Let’s build something amazing together
Let’s build somethingamazing together
For product, interface and web-system work that needs clear direction and careful execution.





