What's the realistic timeline and cost for an agency to rebuild a mid-market storefront on Horizon from an existing Dawn-based theme?
A Dawn to Horizon rebuild costs an estimated $25,000 to $60,000 and runs 8 to 12 weeks for a mid-market storefront (Deploi estimate, illustrative, August 2026). Horizon is a separate theme collection built on theme blocks, introduced 21 May 2025, not a Dawn version update, so theme settings, templates and section content do not transfer (per Shopify's changelog and Help Center, September 2026). Products, metafields and apps carry over untouched.
First, the thing that decides the whole estimate
This is not an upgrade. Shopify's theme update flow applies a new version of the same theme, carries your editor-level changes across, and tells you whether your code edits came with it (per Shopify's Help Center, September 2026). Horizon is a different theme on a different architecture, so none of that applies. You are installing a new theme and rebuilding the storefront inside it.
That sounds worse than it is, because the thing that actually costs money is the section work, and the section work is where the value is.
What the estimate is made of
| Component | Band | Notes |
|---|---|---|
| Discovery, content and section inventory | 1 to 2 weeks | Catalog every section in the Dawn theme and its real usage. Most teams delete 40% here |
| Brand tokens and base setup on Horizon | 1 to 2 weeks | Type scale, color, spacing, block defaults. Do this before any section |
| Core template rebuild | 3 to 5 weeks | Home, collection, product, cart, search, account, plus the content templates |
| Art-directed systems | Parallel, 2 to 4 weeks | Hero and banner system, merchandising rails. See the effort bands below |
| App re-integration and QA | 1 to 2 weeks | Every app embed and app block is re-placed and re-tested |
| Marketing enablement and handover | 1 week | The step that decides whether the build survives contact with the calendar |
| Total | 8 to 12 weeks, an estimated $25,000 to $60,000 | Deploi estimate, illustrative, August 2026 |
What carries over, and what does not
Carries over untouched: products, variants, collections, metafields and metaobjects, customers, orders, discounts, URLs, and your apps as installed. None of that lives in the theme.
Does not carry over: theme settings, section and block configuration, JSON template content, theme code customizations, and any content that was typed into a section rather than stored in a metafield or metaobject. That last one is the expensive surprise, and it is the reason the discovery phase exists.
App blocks are a middle case. They are an Online Store 2.0 and theme-blocks capability, so the app can be placed again in Horizon, but the placement and configuration are manual work per template (per Shopify's Help Center, September 2026).
What moves the number
- How much content lives in section settings rather than in metaobjects. This is the single biggest swing. A storefront whose content is properly modeled rebuilds fast.
- Whether you rebuild the design or port it. A like-for-like port on Horizon is at the bottom of the band. A rebuild with new art direction is at the top, and honestly it should be, because that is a different project wearing the same ticket.
- Number of art-directed systems. Deploi scopes an art-directed hero and banner system at roughly 200 hours and collection merchandising and product rails at roughly 150 hours (directional Deploi estimates from a small sample of engagements, not measured averages). Two systems is a meaningful share of the total.
- App count. Every embed is a re-placement and a regression test.
- Whether marketing is trained. Skipping this saves a week and costs a quarter.
Honest boundaries
This band assumes a single storefront, a single locale, and no replatform of checkout, ERP or search happening in the same window. Multi-market or B2B storefronts sit above it. And there is a legitimate answer that is not on this page at all: staying on Dawn. Dawn is still maintained, at version 16.0.0 released 10 August 2026 (verified September 2026), and "we moved to Horizon" is not by itself a business case. The business case is the section model you build on top of it.
The Deploi point of view
Our own position, from building on Shopify. Separate from the facts above.
- Our take: Do not do this as a theme migration. Do it as a section library build that happens to land on Horizon. The rebuild is unavoidable either way, and the difference between the two framings is whether you own reusable parameters at the end or a second snapshot that will need the same conversation in two years.
- What we’ve seen: The discovery phase reliably deletes a third to a half of the existing sections, and nobody misses them. That deletion is the cheapest cost reduction available in this project, and it only happens if someone is paid to count usage rather than to port the list.
- What it takes: roughly 200 hours for an art-directed hero and banner system and roughly 150 hours for collection merchandising and product rails (directional Deploi estimates from a small sample of engagements, not measured averages).
- Where we disagree: The pitch for Horizon is usually performance and modern architecture. Theme blocks are genuinely better to build on, but no theme swap has ever fixed a merchandising problem, and a like-for-like port to Horizon buys architecture you will not feel for a year.
- What this page adds: that Dawn to Horizon is a rebuild rather than an update, the phase breakdown behind the band, and exactly which storefront data survives the move.
Reviewed by Martin Dejnicki, Director of SEO & AI Search. Facts verified 2026-09-13.
Where we worked this out
Our decision records