How Long Does a Hyvä Project Take?

The honest answer is a range, and it is a wide one. Hyvä puts a custom build at anywhere from 6 weeks to 8 months. An accelerator package compresses that to somewhere between one and eight weeks. Both numbers are real, and the gap between them is not a sales argument, it is a description of two different kinds of work. This article walks through the figures Hyvä publishes, explains why the same technology produces 12 weeks for one merchant and 7 months for another, and lists what actually moves a date once a project starts.

The range Hyvä publishes

Most agencies will not publish timelines. Hyvä does. Thien-Lan Weber, VP of Strategy and Marketing at Hyvä, wrote an article on how long a Hyvä project takes on 30 July 2024, updated since, and the headline figure is that Hyvä projects complete in 6 weeks up to 8 months. What makes it useful is that the article names four real projects behind that range, with team sizes and in one case the hours.

Four projects, four very different answers

PostNL packaging, 12 weeks. A move from Shopware 6 to Magento 2 with Hyvä Theme and Hyvä Checkout. Eight people involved, and a total of 277 hours: 35 of project management and 242 of development, of which 193 came from a single front end developer. That last detail is the interesting one. A twelve week calendar does not mean twelve weeks of continuous work.
Megadental, 7 months, then 1 month. The first project ran 7 months with a technical lead and a junior, at a development cost of 100,000 EUR, including a learning curve of two to three weeks. Their second project took one month. Upgrades and improvements since then take two days a month.
JC-Electronics, 4 months. The full Hyvä suite, Theme, Checkout and Enterprise, integrated by two in-house Magento experts.
Adopt Parfums, 4 months. Building an atomic design system and rebuilding the front end, with four people: two front end developers, one backend developer and a project manager.

Read those four together and the pattern is clear. The shortest project was a platform move with a tight scope. The longest was a first encounter with the technology. The two in the middle were a full suite integration and a design system built from nothing. None of the four was slowed down by Hyvä itself.

Why the same technology gives 12 weeks and 7 months

Hyvä answers this directly, and the answer is experience. Most agencies tell them that their first Hyvä project takes no longer than a Luma project would have. From the second project onwards, they cut development time by up to 50%.

Megadental is the whole argument in one line: 7 months, then 1 month. Same technology, same team, seven times faster. The difference between those two numbers is the learning curve, and somebody always pays for it. On a custom build with an agency new to Hyvä, that somebody is the merchant.

What an accelerator changes

An accelerator package removes the first project from the equation, because somebody already did it. We have written separately on what a Hyvä accelerator package is and who it suits. On Hyvä’s own list of partner packages, the launch times advertised run from one week to eight weeks.

Ours sits at the fast end of that. Liquid Framework goes live in 1 to 2 weeks for a store that fits the standard setup. The qualifier is doing real work in that sentence, and we would rather spell it out than let you find it later: migrations, large catalogs and integrations take longer, and we give the date after the analysis rather than before it.

What actually stretches a timeline

In our experience the theme is almost never the reason a date slips. These are, roughly in order of how often they do it.

Moving an existing store rather than opening a new one. Data, URLs, customer accounts and order history all have to arrive intact, and the old store has to keep selling while they do. That is a different question from this one, and we answer it in our guide to migrating a live store to Hyvä.
Integrations with systems you already run. An ERP, a PIM or a WMS sets the pace, because the other side has its own release calendar and its own people. This is the single most common reason a two week launch becomes a two month one.
Catalog size and catalog quality. Thousands of products are not a problem. Thousands of products with inconsistent attributes, missing images and three generations of category structure are.
A design system built from nothing. Adopt Parfums spent four months on exactly this. If your brand has a design language that must be followed precisely, budget for it as its own piece of work.
Features that only Adobe Commerce ships. The B2B suite, customer segments and content staging are not in Magento Open Source. On Open Source they get built, and B2B in particular carries surprises.
Third party modules that have to be rewritten. Anything that rendered its own interface on the old theme needs a Hyvä version. Counting those early is the difference between a confident date and an apologetic one.
Contracts you have to sign yourself. Payment providers and couriers contract with you, not with your agency. Those approvals sit outside anybody’s development plan and they are a common reason a finished store waits.
Decisions and content on your side. Copy, photography, legal pages and the person who signs off. A build waiting for approval looks exactly like a late project on a calendar.

What a one to eight week window is actually made of

If a launch window is measured in weeks rather than months, it helps to know what fills them, because the work does not vanish. It moves earlier, and a good deal of it moves to you.

Analysis. Somebody reads your catalog, your current modules, your integrations and your traffic, and produces a date from them. On an accelerator this is the only stage that can still move the date materially, which is exactly why it belongs before the proposal rather than after it.
Configuration rather than construction. The theme, the module set and the checkout already exist. What happens in these weeks is fitting them to your catalog, your shipping rules, your tax setup and your brand. Nobody is writing them again.
Content. Category copy, product data, photography, legal pages. This is the stage that most often sits waiting, because it is the stage nobody can do on your behalf.
Payment and courier approvals. Started in parallel on the first day, because they run on somebody else's calendar and will not be hurried by yours.
Testing, then go live. Real devices, a real card, a real order through every payment and delivery method you intend to offer.

Read that list and the honest reason an accelerator launch is fast becomes obvious. It is not that the work is smaller. It is that most of the technical work was done once, for everybody, before your project started. What remains is the part that was always specific to you, and that part is mostly decisions. Which is also why the merchants who launch in two weeks are, almost without exception, the ones who answered their emails.

How to read a date you are given

A promised launch date is only worth what the assumptions under it are worth. Four questions tell you whether a number is real.

Is this date before or after the analysis? A date given before anyone has seen your catalog and your integrations is a hope, not an estimate.
How many Hyvä projects has this team shipped? Megadental shows what the first one costs, and it is not the agency that pays.
What does the date assume about me? List what you owe them: content, approvals, access, provider contracts. Then be honest about whether you can deliver it.
What is explicitly outside the date? Integrations and migrations usually are. Better to read it in a proposal than to discover it in week six.

The next step

If you want a date you can plan around rather than a range, the analysis is what produces it. You can see what we have shipped first if you would rather start there. When you are ready, book 30 minutes and we will go through your catalog, your integrations and what you already run, and tell you which of the two routes your store is actually on.