Short answer
A Hyvä migration costs £8,000 to £80,000+ depending on store size and third-party module complexity, with most mid-market Magento stores landing between £15,000 and £35,000 for a 3-to-5-week project. The Hyvä Themes license itself is £1,000 to £3,000 one-time per domain. The rest is engineering time. Industry-typical ranges from agencies actively shipping Hyvä work as of May 2026, against retail and B2B engagements we have run or benchmarked since 2021.
What you are actually paying for
A Hyvä migration is not a re-platform. The Magento backend stays where it is. What you are paying for is engineering time on three distinct phases:
- Third-party module audit and reconciliation. The biggest variance driver. Each non-Hyvä module needs a compatibility variant, a community port, an in-house adapter, or a removal decision.
- Theme implementation. Branding, design translation from Luma to Hyvä, custom blocks, page-builder content adaptation.
- Testing, cutover, post-launch monitoring. Staging dry run, off-hours production cutover, 72-hour monitoring, regression fixes that surface in week one.
Cost by store size
Small Magento store: £8,000 to £15,000
Single store, under 1,000 SKU, simple checkout flow, fewer than 10 third-party modules, no heavily custom theme overrides. Most agencies will fixed-price these. Timeline: 2 to 3 weeks. The Hyvä Themes license sits at the lower end of its £1,000 to £3,000 range. Engineering time accounts for the rest. Most of the budget goes to audit, theming, and cutover testing rather than third-party reconciliation.
Mid-market Magento store: £15,000 to £35,000
1,000 to 50,000 SKU, B2C, 15 to 25 third-party modules, custom Knockout components, multi-language or multi-currency. Timeline: 3 to 5 weeks. The Hyvä Themes license is typically at the higher end. The variance inside the £15-35k range is almost entirely driven by how many third-party modules need compatibility work. A store with 20 modules where 18 have official Hyvä variants costs much less than the same store where 12 need community ports or in-house adapters.
B2B or enterprise Magento store: £35,000 to £80,000+
Complex catalogs (50,000+ SKU), customer-group pricing, quote workflows, multi-store, multi-currency, multi-warehouse, 30+ third-party modules including B2B-specific extensions. Timeline: 5 to 10 weeks. Adobe Commerce stores often land here because of the additional Page Builder content adaptation work. Fixed price becomes harder; many agencies (including us) move to a capped time-and-materials structure on engagements this size.
What drives variance inside each tier
- Third-party module count and Hyvä-readiness. The single biggest variance driver. Audit before you scope.
- Custom Knockout components. Anything in
view/frontend/web/jsneeds an Alpine equivalent. Easy to under-scope. - Page Builder content blocks. Adobe Commerce stores with heavy Page Builder content need template overrides per block type.
- Brand complexity. A clean modern brand translates cheaply. A heavily-illustrated, multi-typeface, custom-animated brand costs more.
- Multi-store / multi-language. Each store view multiplies QA effort. Two store views ≠ 2x cost, but 5 store views can mean 30 to 50 percent more testing time.
- Geography and agency tier. UK and EU agencies typically charge £80 to £150 per engineering hour. US agencies often charge more. Offshore agencies advertise lower rates but quality variance is significant; project length often offsets the rate difference.
Hidden costs to watch for
- Hyvä compatibility module renewals. Some compatibility modules carry annual renewals. Budget for them year-on-year.
- Hyvä Checkout licensed separately. Not included in the Hyvä Themes license. Add £1,000 to £3,000 if you adopt it during the migration (most stores do).
- Analytics rewiring. Knockout-based event handlers for GA4, Meta Pixel, and ESPs need Alpine equivalents. Usually 1-3 days of engineering not in the base scope.
- SEO testing post-launch. Crawl differences between Luma and Hyvä can surface unexpected canonical or robots issues. Plan a week of Search Console monitoring.
- Email template adaptation. Transactional emails (order confirmation, shipping) need the same brand translation. Often forgotten in the scope.
Fixed price vs time-and-materials
Fixed price is preferred for small-to-mid-market migrations where the third-party module audit produces a clear scope. The agency carries the risk of underestimation; you carry the risk of needing changes outside scope (paid as variations).
Time-and-materials (T&M) with a cap is common for B2B and enterprise migrations where the scope cannot be fully fixed upfront. You pay weekly or fortnightly against a cap; the agency stops billing at the cap regardless of remaining work. Transparency on time tracking is the load-bearing requirement.
Monthly retainer billing for a Hyvä migration is a red flag. Migration is project work. A retainer hides scope drift and makes it impossible to assess true cost.
DIY vs agency cost comparison
An in-house engineering team can run a Hyvä migration. The cost is internal engineering time, typically 3 to 8 engineer-weeks for a mid-market store, against an internal blended cost of £80 to £200 per engineer-hour. That works out to £8,000 to £40,000 of internal cost, which is the same order of magnitude as an agency.
The DIY argument is strongest when the team already knows Hyvä and Tailwind, or when the engineering cost was already allocated and Hyvä displaces other work. Agency engagements win on time-to-launch and on the breadth of past Hyvä mistakes already debugged.
What "too cheap" and "too expensive" look like
Below £8,000 for a mid-market store: likely a junior-team scope that does not include third-party reconciliation, or an offshore quote where revisions and delays will inflate the real cost.
Above £80,000 for a B2C store under 50,000 SKU: likely a large agency adding overhead, or scope creep absorbing project budget. Get a second quote.
Subscription pricing for a migration project: red flag. Migration is finite work. Subscription disguises scope and runs longer than necessary.
Frequently asked questions
How long should a Hyvä migration take?
Two to ten weeks depending on store size. Most mid-market engagements land between three and five weeks of wall-clock time. Anything proposed under two weeks for a real store is usually a partial migration that leaves significant work for you to finish after the agency walks away.
Should we migrate during peak season?
No. Plan the cutover for a low-traffic window. UK retail: a Tuesday evening in February tends to be the calmest. US: avoid the September-to-January window. UAE: avoid Ramadan and Eid-adjacent periods.
Does the Hyvä license transfer if we change agencies?
Yes. The Hyvä Themes license is owned by your store, not by the implementing agency. Make sure the license is registered to your company name, not to the agency's account, at purchase time.
Can we phase the migration to spread cost?
Yes. Stage 1 ships home and category templates on Hyvä, leaving product and checkout on Luma. Stage 2 migrates the rest. The performance gain is smaller mid-phase but the cost is split across two budgets. We have shipped this pattern when CFO timing constrained the engagement.
What's included in an agency Hyvä migration quote?
Standard inclusions: third-party module audit, Hyvä Themes license (sometimes), implementation, branding, multi-store configuration if applicable, staging deployment, production cutover, 30 days of post-launch support. Always-check exclusions: Hyvä Checkout license, custom Page Builder adaptation, transactional email rebuild, analytics rewiring, SEO monitoring. Confirm these line by line.
Next step
Run the migration math for your store. Inventory your composer.json third-party modules against the Hyvä Themes compatibility list to estimate your reconciliation burden. That single number, plus your store size, predicts most of the cost variance.
The full Magento to Hyvä migration guide covers the project plan in detail. If you would like a fixed quote, the Hyvä migration service scopes within one business day. Initial scoping call is free.