# How Kiwi tried to fix Ethereum mobile UX with MWP

Source: https://x.com/MacBudkowski/status/1920137264912318707?s=19&t=WYqrKlGfxEelMGR0daBqaQ
Kiwi News: https://news.kiwistand.com/stories/How-Kiwi-tried-to-fix-Ethereum-mobile-UX-with-MWP?index=0x681b7d0f213b25fb6c6ec2de5df7d5189f84d6601a38f7813c4fcc381ed2045494c4b701
Submitted by macbudkowski.eth on 2025-05-07 UTC, 2 upvotes, 0 comments

## Summary

The Kiwi team argues that Ethereum mobile wallet connections are unreliable and that Mobile Wallet Protocol (MWP), built by Coinbase, offers a fix. They say WalletConnect connections work only about 80% of the time on mobile, while MWP works about 99% of the time and lets iOS users be redirected back to the dapp after connecting. Kiwi shipped MWP in its iOS app, though the protocol does not support web apps and currently works only with CB Wallet and Rainbow. The team also calls on the Ethereum ecosystem, including the EF, to give mobile UX more attention.

## Article

"We should be as decentralized and open as Linux and Android, and at the same time have as consistent user experience as Apple"

- Vitalik once said.

But if Steve Jobs saw Ethereum mobile UX, he'd tell us to get the hell out of his office.

Good news is that we might have found a silver lining that could at least partially fix Ethereum mobile wallets UX.

Here's a story:

For the last 2 years we've been trying to onboard users to our mobile dapp at many ETH conferences around the world, from Bangkok to LA.

We quickly noticed that mobile wallet connections work maybe 80% of the time. Which is fine for a hackathon project, but embarassing for a network that wants to onboard billions of users.

Can you imagine Instagram, Tinder or Uber founders trying to onboard users while 20% of the time these users can't even log in? That would be an absurd growth blocker. But it's been an every day life of Ethereum dapp builders.

And that despite that mobile is where our future userbase is: Africa, Asia, Gen Z. These are billions of people that need to be served well.

But who cares, right?

Ethereum is full of laptop-first millenials like me who are stuck to desktop wallets. They don't even experience the pains of using Ethereum on their phone, so they don't discuss it here.

And there are many pains:

(1) Wallets load too long.

Last time Rainbow took 10 seconds to show me the "Connect" button. Users in their 'mobile attention span mode' can't stand waiting so long.

(2) Broken connections.

Sometimes the in-wallet "Connect" notification never appears. So the user thinks that your app is broken and leaves it.

(3) In-wallet "Connect" button doesn't work.

Sometimes the user clicks "Connect" in their wallet, and nothing happens. The users will tell you "Ah yeah, it happens sometimes. I will try again at home", and then they get distracted and never open your app again.

(4) You don't get redirected back to the app.

A user clicks "Connect" in their wallet, but they need to manually open the browser or app again. An average user gets confused, distracted and never goes back to your dapp.

Building for consumers is hard even without all these problems.

Consumers are used to slick web2 UX. Their phones are bombarded with notifications from other apps. And whenever people are on their phones, their tempers become shorter and more sensitive to bugs.

So it sucks that Ethereum dapp builders have this extra thing to solve.

Why it sucks is a bigger question. These problems lie at the intersection of WalletConnect, Ethereum mobile wallets, dapps, and wallet connectors. And probably that's why these problems haven't been solved yet.

@timdaub and I spent 2 years complaining about it, but CT is more interested in minor details of another Ethereum fork than actually improving mission-critical UX. So we did our own research.

It took many DMs and a lot of digging, and we found that - buried under the rock - there is a solution that might improve Ethereum UX on smartphones.

It's called Mobile Wallet Protocol (MWP), built by Coinbase. If you used the Warpcast mobile app, you might have already experienced how it works under the hood.

It uses deeplinking, is specialized in native mobile apps and is way more reliable than WalletConnect there. This makes MWP work about 99% of the time, like a solid consumer-grade technology.

So once we discovered it, we started chatting with other dapp builders and created a group chat for devs to swap notes and discuss how to implement the MWP. It wasn’t easy, because we had to figure out a lot of things from scratch.

In the meantime, we found that MWP lets us achieve the Holy Grail of iOS - once a user connects the wallet, we can redirect them back to the dapp (as you see in the video). It has been possible on Android (and is doable via WalletConnect), but on iOS it was hard to do.

After a few weeks of testing, we finally shipped MWP in our Kiwi iOS app, and it 'just works'. Like core infra should.

And yes, MWP is not a perfect WalletConnect replacement. For example it doesn’t support web apps, and at the moment only works with CB Wallet and Rainbow. 

But there’s a lot of potential to grow and improve it and we hope that @jessepollak will follow through with this protocol. And maybe some competition would motivate the WalletConnect team to push forward even more.

BTW it is pretty sad that talking about problems with mobile UX for 2 years didn't result in any reactions from the Ethereum ecosystem. 

That dapp builders are left to their own devices and have to search for ways to fix these problems on groupchats and informal channels.

We hope, though, that learning about these mobile UX problems and seeing some solutions will inspire our ecosystem to give Ethereum mobile more love. 

Especially since we have people like @marek_ and the Celo team that knows a thing or two about building mobile-first blockchains, and could share their lessons learned.

It is not an easy fix, and takes a lot of coordination, so it would be great if @tkstanczak and the EF helped to lead the ecosystem to fix it.

Now is a critical moment to act because if we don't improve our smartphone experience, we will lose ground to mobile-optimized networks like Solana. 

Whatever happens, we hope we can finally start onboarding users like it’s actually 2025.

Because how could we onboard the next billion users without really good mobile experience?

PS: thanks to @FUCORY, @jamalavedra, @coinspect, & @glitch_txs for helping us figure things out