Should you migrate to Shopify Horizon?
The honest version — including the cases where the answer is no.
We migrate Shopify stores to Horizon for a living, so read this with that in mind. It’s also why we’d rather tell you when not to do it than sell you a rebuild you didn’t need.
No email required to read this page. The full framework, with the app compatibility reference and a scoring worksheet, is at the bottom.
Four things worth knowing before anyone sells you anything.
No deprecation has been announced. Dawn still ships, still receives maintenance updates, and Shopify has given no indication it’s going away. Anyone telling you there’s a deadline is inventing one.
Public theme store sentiment sits at roughly a third positive. The recurring issues are theme updates overwriting merchant customizations, and mobile bugs following new releases. Both are worth understanding before you commit — we cover the first one below, because it’s largely a consequence of how a theme gets built rather than the theme itself.
On any store that’s been customized, speed is mostly determined by what’s been added to it — apps, third-party scripts, tag managers, image handling, marketing tags — not by which theme it started from. Theme-versus-theme benchmarks compare vanilla installations, which is not a store anybody actually runs. A rebuild is a genuine opportunity to fix performance, because you get to leave the accumulated weight behind. It is not a guarantee of it, unless somebody puts one in writing.
Horizon reconstructed the variant picker, gallery, cart drawer, and predictive search as self-contained web components. Themes don’t port. Custom sections don’t carry over. Some apps written against older markup will break.
None of that is an argument against migrating. It’s the set of facts any honest recommendation has to start from.
Because the constraint most stores actually have isn’t speed. It’s throughput.
On a mature store with years of custom sections, routine merchandising work still needs a developer. A new campaign page. A layout change. A promotional module before a launch. None of it is technically hard, and all of it goes into a queue behind work that genuinely is.
That’s the problem Horizon’s architecture addresses. Nested, reusable blocks mean layouts get assembled rather than built. AI-generated theme blocks mean a described change can become a real one without a ticket. Structure and layout control moves to the people who want the change made.
The second reason is where Shopify is spending. Storefront capability is being built on the newer architecture, and the investment gap is visible in the public record:
| Dawn | Horizon | |
|---|---|---|
| Releases, last 12 months | 3 | 19 — roughly one every three weeks |
| Major versions, last 24 months | 1 | 4 |
| Most recent major version | v16.0.0, August 2026 | v4.0.0, June 2026 |
| Largest single release, last 12 months | v15.5.0 — 112 files changed | v4.0.0 — 314 files changed |
A single Horizon release in December 2025 changed more files than everything Dawn has shipped since, combined.
The nuance most coverage misses: Dawn still gets the platform plumbing it needs to keep working. Its v15.5.0 release added support for standard storefront events and for app, agent, and AI cart interactions without page reloads — so Dawn isn’t frozen out of agentic commerce — and its v16.0.0 major version in August 2026 moved the theme onto Shopify’s customer accounts. Dawn gets the plumbing. Horizon gets the product.
Migrate now if two or more of these describe you.
This is the clearest case. You’re rebuilding anyway, so the only real question is which architecture you rebuild onto. Doing it twice costs far more than doing it once.
If campaign pages, layout changes, and promotional modules routinely wait on developers — and the wait is measured in weeks — you’re paying for that constraint every quarter, whether or not it appears anywhere in a budget.
An agency you no longer work with built it. The person who knew it left. Every change is now archaeology. Even if you decide against migrating, this is a problem that gets more expensive with time.
Complexity that’s merely annoying at one storefront becomes structural at four. Every change made four times is the definition of a throughput problem.
Subscriptions, reviews, CRO tooling. Theme work is adjacent to all of it, and doing them in separate windows means doing overlapping work twice.
If you can no longer safely take upstream updates because too many core files were modified, you’re already carrying the cost of a migration — you just haven’t scheduled it.
Wait if any of these describe you.
We give this recommendation two or three times a quarter. It isn’t a hedge.
If you’re on a modern, well-maintained theme, your team can already make most changes without a developer, and nothing is blocked — the case for a rebuild is weak. Revisit when a redesign appears on the roadmap.
If the subscription platform, bundling engine, or personalization tool your revenue depends on has no Horizon-compatible version and none on its roadmap, migrating means rebuilding that functionality yourself or losing it. Wait for the vendor.
There’s a threshold where accumulated complexity makes the rebuild larger than the problem it solves. It’s not common, but it’s real, and it’s worth measuring before committing.
If you’ve recently invested in a modern theme build that’s working, the return on doing it again is low. The exception is if that rebuild is already showing the same constraint — a new theme that still routes every change through a developer hasn’t solved the problem you have.
A migration needs a day or two of real attention from someone who knows how the store works. Without that, the specification is guesswork and the result will disappoint you regardless of who builds it.
Five questions. Your answers will tell you more than any article will.
Under a day means your setup is working. Two weeks or more means the architecture is the constraint, not your team.
Not the things you decided against. The things you wanted and couldn’t. That list is the actual cost of your current setup, and most teams have never written it down.
If the answer is one person, or an agency you no longer use, or nobody — that’s a risk you’re carrying whether or not you migrate.
If the answer is “we can’t” or “it broke things,” your theme has drifted from its base and you’re already accumulating migration debt.
A redesign, a new market, a replatform of part of the app stack — any of those changes the calculation, because they mean the work is happening anyway.
Nobody publishes this, so here’s what’s actually out there.
There is no reliable public benchmark specifically for Horizon migration pricing. What exists are broader Shopify agency ranges, which are the closest useful reference:
| Type of work | Typical range |
|---|---|
| Theme customization or small features | $5,000 – $15,000 |
| Full-site design and build | $25,000 – $80,000 |
| Agency-assisted mid-market migration | $15,000 – $75,000 |
| Moderately complex mid-market migration | $45,000 – $75,000 |
| Custom Shopify Plus replatform | $75,000 – $250,000+ |
Sources: published US agency data and 2026 market guides.
The distinction that matters more than the number: theme customization and migration are different categories of work, and quotes at the bottom of that range are almost always the former. If you’re comparing a $12,000 quote against a $45,000 one, check whether they’re describing the same job before concluding one is overpriced.
The other question to ask any quote: what happens when they find something in week three that wasn’t in the estimate. A number that can move isn’t really a number.
Three things that determine whether it goes well.
The most common failure in a theme migration is functionality that quietly disappears because nobody wrote it down. Ask any agency how they capture what your current theme does. If the answer is “we’ll go through it with you,” they’re relying on your memory of four years of changes, which is not a plan.
This is the single most common complaint about Horizon, and it’s mostly avoidable. Custom work belongs in separate blocks and section files, never in core theme files. A shop that edits core files will hand you something that breaks on the next update.
Every app in your stack should be scored against Horizon — works as-is, needs reconfiguration, needs a vendor update, or needs a rebuild — with a cost attached to each red item. Finding this out during the build is how fixed prices become change orders.
Get the full version.
The complete framework as a PDF:
- The scoring worksheet — four dimensions, with the rubric, so you can score your own store
- The app compatibility reference — where the common DTC apps currently stand on Horizon
- A pre-migration checklist covering requirements, SEO, tagging, and rollback
- The questions to ask any agency before signing
Email only. No call unless you ask for one.
Common questions.
Is Shopify Dawn being discontinued?
No. Shopify has not announced any deprecation of Dawn. It still ships, still receives maintenance updates, and there is no published end-of-life date. What has changed is where new storefront capability is being built.
What’s the difference between Dawn and Horizon?
Dawn is an Online Store 2.0 theme built on sections with limited block nesting. Horizon uses a block-based architecture with far deeper nesting, self-contained web components for elements like the variant picker and cart drawer, and AI-assisted block generation in the editor. In practice the difference is less about appearance and more about how much your team can change without a developer.
Will migrating to Horizon make my site faster?
Not on its own. Once a store has been customized, its speed is driven overwhelmingly by what’s been added — apps, third-party scripts, tag managers, image weight — rather than by the underlying theme. Published theme comparisons measure vanilla installations, which nobody runs. What a migration does give you is the chance to rebuild without carrying the accumulated weight forward, which is where the real gains come from. Treat performance as something to be measured and committed to for your specific store, not something a theme choice delivers by itself.
Can I just switch my theme to Horizon?
No. Themes aren’t portable — your custom sections, settings, and integrations don’t carry across. Moving to Horizon means rebuilding your storefront’s functionality on a different architecture.
Will my apps still work on Horizon?
Many will. Apps using theme app extensions generally port cleanly. Apps that were hardcoded into your theme, or written against Dawn’s markup, frequently break — Horizon rebuilt several core elements as web components. Check with each vendor, or have the stack mapped before you commit.
How long does a Horizon migration take?
It varies with complexity. A single-store, single-market migration is typically a few weeks of build time once requirements are documented. Multistore or multimarket work takes considerably longer.
Can I migrate to Horizon myself?
If your store is relatively simple and your team has the capacity, yes. The work that catches people out isn’t the build — it’s documenting what the current theme does, mapping app compatibility, and preserving SEO and tracking. Those are the parts worth being systematic about regardless of who does the building.
Do I need to be on Shopify Plus?
No. Horizon is available across plans, though some features — combined listings, for example — are Plus-only.
What happens to my SEO if I migrate?
Nothing, if it’s done properly. URL structure should be preserved, existing redirects carried over, and structured data rebuilt to match what you currently emit. It’s damaged when those are treated as afterthoughts, which is common, because nothing looks broken on launch day — the traffic loss shows up weeks later.
Should I wait for a later version of Horizon?
If you have no redesign planned and nothing is blocked today, waiting is reasonable. If you’re rebuilding anyway or your team is stuck behind a dev queue, waiting means a bigger rebuild later and another cycle of the same constraint.
Still not sure?
The fastest way to make this concrete is to see where your store actually stands — speed, accessibility, technical SEO, and theme architecture, measured against three competitors you pick.
Two minutes. Nothing to install. We’ll send it back within a day.
If you’d rather go further, our Readiness Audit is a five-day assessment that answers this question specifically for your store — including whether the answer is no.
