I work with Magento and Hyvä every day, and store owners keep asking me the same question: should we move to Hyvä now, or wait a bit longer? I understand the caution. The store makes money, and any move comes with the risk that something stops working at the worst possible moment.
So let’s set aside agency opinions for a moment, ours included, and look at the data.
What every second costs you
Start with a customer opening your store on their phone. Google put a number on it back in 2017: 53% of people leave a mobile page that takes longer than three seconds to load1. That data is getting on a bit, but newer measurements point the same way.
The agency Portent analysed more than 100 million page views and looked at how store conversion rates change with page load time2:
Look at the gap between the first row and the last. In this study, a store that loaded in one second sold more than four times as often as one that loaded in four.
Google and Deloitte measured it differently, across 37 brands and more than 30 million sessions. Making a mobile site a tenth of a second faster was associated with an 8.4% higher conversion rate in retail, and shoppers spent 9.2% more3. Vodafone tested it with an A/B test: the version of the page whose main content loaded 31% faster sold 8% more4.
And what about Poland, where we work? According to Gemius, 58% of Polish internet users finish their purchases on a phone, and stores see on average 14 million daily users on smartphones against about 3 million on computers5. A store that is slow on phones is slow for most of its customers.
What Hyvä actually is
Magento has two layers. The first is the back end: products, prices, orders, customers and the connections to your ERP or warehouse. The second is the frontend, meaning everything the customer sees in the browser. Magento’s default frontend is called Luma, and it loads a lot of code that the phone has to download and run before it shows the page.
Hyvä replaces only that second layer. Products, orders, page addresses and integrations stay where they were, so you don’t migrate any data6. The new frontend is written from scratch with lighter tools, which is why it loads faster.
Hyvä is made by Hyvä Themes, a company that launched in 2020. For years a theme licence cost about €1,000 per store7. On 10 November 2025 the theme became free and open source, under the same licences as Magento Open Source8. The add-ons are still paid, and for a store owner the one that matters most is Hyvä Checkout, the new checkout process: it costs €1,000 one-off, with a year of updates and support included9.
With the licence off the table, what a migration costs you is developer time.
What the data from thousands of stores says
Hyvä shows success stories on its website, but the vendor always picks those. The more interesting data is the kind nobody picked. Every month HTTP Archive measures real visits to websites, including more than 32,000 Magento stores and almost 4,700 on Hyvä10.
The yardstick is Core Web Vitals, Google’s three metrics: how quickly the main content appears, how quickly the page responds to a tap or click, and whether elements jump around while loading11. Google uses them in the systems that decide the order of search results. It adds that good scores don’t guarantee a high position, because relevance matters most12.
Hyvä stores also load less. The median amount of JavaScript per page is 767 KB against 1.27 MB across Magento, and the median page weight is 2.25 MB against 3.39 MB10.
One number in this data cools the enthusiasm. Server response time is good for 23.2% of Hyvä stores and 23% of Magento stores, practically the same10. A new frontend won’t fix a slow server or an overloaded search engine. If the store is slow on the back end, that needs work too.
In October Store Leads counted 7,134 stores on Hyvä, 22.7% more than a year earlier. Poland, with 256 stores, ranks eighth in the world13.
Success stories, and the small print
The Hyvä website has some impressive results. The Swiss chocolatier Läderach reported 47.8% more conversions and 39% more revenue after moving to Hyvä14. The British Citizen Watch store reports a 43% higher conversion rate and 141% more revenue15.
Those numbers are real, but you need to know where they came from. Hyvä publishes both write-ups, and in both cases the store also got a completely new design14,15. Part of the growth comes from speed, part from the new design, and part can’t be put down to either, because seasons and campaigns change in retail too.
So when you judge a migration, look first at the store’s speed before and after. Compare sales in your own data, keeping the season and your campaigns in mind.
Where a migration can hurt
Think of your store as a house where you’re replacing the whole facade and the windows. The structure stays, but everything bolted to the outside has to come off and go back on. In a store, those things are the extensions.
Extensions. Modules written for Luma need a compatibility module that rewrites the part customers see16. Hyvä keeps a public list of these modules: it tracks 968 extensions, 526 of them marked as compatible17. If your store uses a rare module or code added by a previous agency, its frontend will have to be rewritten.
Checkout. At first the checkout can keep Luma’s old look while the rest of the store already runs on Hyvä18. That’s handy as a start, but it means the store maintains two different frontends for a while, and every change to the basket is done the old way. When you move to Hyvä Checkout, you also need to check that it supports your payment and delivery methods6.
Scope. The British agency iWeb warns against treating a migration as an ordinary theme swap:
Treating a large trade estate as a simple theme swap can stretch a planned fortnight into a quarter. iWeb, Luma to Hyvä migration service page
The US agency IWD quotes 8 to 12 weeks and $35,000 to $60,000 for a store with 10–30 extensions, and up to 20 weeks and $150,000 for stores with more than 50 extensions and a custom checkout19. Those are US figures, so treat them as an order of magnitude, not a quote.
There’s also an argument for not waiting forever. Adobe doesn’t support Luma in its new cloud version of Adobe Commerce, and it recommends its own frontend built on a different technology for new stores20. Existing Luma stores keep working, but you can see which way the vendor is heading.
How it went for us
We work with Hyvä in several stores. The numbers below come from our projects:
- Zegarki Błonie. We built a new Hyvä theme for the whole store in about two weeks. The home page now loads in about 2 seconds instead of 4, and the store kept its look: the handwritten logo, the colourful backgrounds and the large watch photos. See the project.
- Quiosque. We took over this store of more than 5,000 styles from a previous agency, removed the bottlenecks in its Hyvä frontend and rolled out Hyvä Checkout without pausing orders. On phones the store scored more than 85 points in Lighthouse, despite large photos from fashion shoots. See the project.
- System 3D. Here we first removed the causes of outages: we updated Magento, tidied up the catalogue and set up Varnish. Only once the store was stable did we start moving it from Luma to Hyvä and Hyvä Checkout. That work is ongoing. See the project.
- .mdd. Working alongside this furniture maker’s in-house team, we helped move the store to Hyvä. After the migration the frontend runs five times faster. See the project.
System 3D shows the order I described above. If the store went down because of its search engine, a new frontend wouldn’t have changed that.
Is now the right time for your store?
These are the questions we start a migration conversation with:
- Measure your store on a phone. If the main content of your home page, a category or a product page appears later than 2.5 seconds, Google rates that as needing improvement11.
- Find where the problem is. A slow server or an overloaded search engine will stay slow on Hyvä too. Those get fixed separately, ideally before the migration.
- List your extensions. For each one, check whether it has a Hyvä compatibility module. The ones without push up the cost.
- Decide what happens to the checkout. Hyvä Checkout straight away or in a second stage. Check that it handles your payment and delivery methods.
- Plan whether you're changing the design. A migration is a good moment to refresh the store, but it then gets harder to tell what the new frontend alone delivered.
How we run a migration
First we look at the store: the Magento version, the server, the extensions and the custom code. That tells us how much work the frontend needs and what has to be fixed on the back end. Then we migrate from Luma in stages, without interrupting sales, and the brand’s look and the store’s data stay in place. You can read more about how we work with this theme on the Hyvä and Hyvä Checkout pages.
If you remember one thing from this post, let it be this: Hyvä speeds up what the customer sees, and the licence no longer costs anything. The stores that gain most from a migration are the ones with a healthy server and a list of their extensions.
