Shopify Theme Trends 2026 [85,445-Store Study]

TL;DR: StoreInspect’s retained April 17, 2026 study compared endpoints for 85,445 rescanned domains within a 120-day lookback. The broad Custom classification rose 262 (1.4%); literal Custom name rows rose 468 (35.7%). Horizon had the largest absolute gain among named products in the selected gainer table, +84, while Tinker’s 20.6% relative change exceeded Horizon’s 15.2%.

Some links in this article are affiliate links. We may earn a commission if you purchase through them, at no extra cost to you.

A static theme popularity study describes recorded size. Repeated snapshots can also reveal changes in public theme labels. This article describes historical endpoint changes; it does not verify redesign projects, bespoke builds, purchasing budgets or October theme trends.

How the lookback panels were selected

The retained shopify-theme-trends-data-summary.md reports 634,876 snapshots across 534,514 domains. Overall database timestamps range from December 8, 2025, 07:08:37.296 to April 17, 2026, 07:09:10.597; the summary does not specify their timezone or record an exact execution-time cutoff.

For each lookback, the query keeps snapshots newer than execution time minus 30/60/90/120 days, requires a non-null theme and a name other than Unknown, then requires at least two observations per domain. It compares the first and latest eligible observations within that lookback. The count rule does not require a full 120-day elapsed span—or even distinct observation dates—for every domain.

These are different selected cohorts with potentially overlapping domains and different endpoints. Their rates do not isolate the effect of observing the same merchants longer. Per-domain span/age distributions and exact intersections were not retained. The overview counts name or type changes, so a classification-only change can qualify even without a new name.

Public theme signals and catalog mapping can change. Merchant renaming, heavily modified themes and detector/classification changes can appear as transitions. No stable-version or manual transition-validation sample accompanies this extraction, and snapshot presence does not independently verify trading activity.

For current single-store checking, use the theme detection guide. For measured performance and its coverage limits, see theme performance; label changes alone establish no speed or conversion effect.

Broad buckets and literal theme names are different

The bucket query first assigns selected legacy names case-insensitively, then selected modern free names, then Custom if the name is custom or the stored type is custom, and Paid by stored type. Other/unclassified names are outside the four displayed buckets, so these rows do not sum to the full panel.

The separately grouped literal Custom name goes from 1,310 to 1,778 (+468, 35.7%). That is not the broad Custom bucket’s 18,951 → 19,213 (+262, 1.4%). A commercial theme, modified base or unresolved public identity can receive a custom classification; neither number establishes bespoke development or brands “graduating” from paid themes.

The legacy bucket’s case-insensitive rule uses Debut, Brooklyn, Minimal, Simple, Supply, Venture, Boundless and Narrative. A separate case-sensitive exact-name legacy query reports 10,656 → 10,541, with 129 exits and 14 entries. The different casing predicates explain why the starting levels cannot be interchanged. Both saved outputs show −115, but neither establishes architecture/version, security support, motive or current need for redesign.

Selected named-theme gains and losses

The net gainer/decliner queries require at least 100 matches at either endpoint, order by absolute net change and retain 20 rows each. The following displays selected rows; it is not a complete theme census or a percentage-growth ranking.

Horizon’s 84 net gain is the largest named-product absolute increase in that selected result. Tinker’s smaller starting base yields a higher percentage change, 20.6% versus 15.2%. In a separate arrival query, 97 domains changed from another name to Horizon, representing 15.2% of 638 latest Horizon matches. That arrival share and the coincidentally equal net-growth percentage are different measures.

A falling recorded name count can reflect changed usage, naming or detection. It is not proof of declining theme quality, unsupported architecture or a merchant wanting an agency.

Direct endpoint label paths

The retained direct-path query compares name changes, requires at least five matching domains per path and retains the top 20. These selected rows therefore omit smaller transitions and do not capture intermediate themes used between endpoints.

These paths end at the literal name Custom, not the broad Custom classification. Dawn → Custom appears on 72 domains; that does not establish successful scaling, demand proof or a specific custom build. Reverse movement appears too, including Custom → Dawn on 13 domains.

The Horizon source table is separately filtered to at least two domains per source and retains 15 rows:

Dawn supplies the largest listed source count, 20. Confirm the underlying theme identity and a real merchant project before describing a transition as an upgrade or using it to time outreach.

Segments show recorded differences, not budgets

The estimated 50K–200K group accounts for 1,080 name changes. The 200K–1M group has a higher recorded rate, 5.64%, on a smaller base. The tiny 1M+ buckets do not support a general high-traffic pattern. These are associations in selected rescans, not evidence of revenue or redesign spend.

The query groups a true stored Plus flag separately and folds false or null into its other group. It is a detection classification, not billing-confirmed plan status. Its changes sum to 1,699, while the overview reports 1,703: the segment query tests name only, whereas the overview tests name or type. Separate historical statements also were not a frozen transaction, so the four-record difference should not be used to reconstruct missing observations or unknown flags.

The rates are about 1.9 times apart, but no budget, merchant motive or adjacent-app purchase was measured. The literal Custom arrivals—438 versus 95—likewise do not prove custom-project spending.

Named-category observations

The category query requires at least 500 matched domains and retains 12 rows. Other contains 51,525 domains and 994 name changes, so the named categories below describe only part of the panel.

Fashion has the most changes among the displayed named categories, 198; Beauty has 151 and a higher rate, 2.85%. These counts do not establish that visual storytelling drove transitions or that the merchants need premium services.

Custom arrivals here again mean the literal name. Category assignment, theme labeling and selective rescans affect these comparisons; a catalog-heavy category is not evidence that a free theme has been outgrown.

Use a label transition to choose a verification step

An agency can check whether the public storefront changed and whether the merchant has a concrete unresolved design task. A theme developer can inspect how a supported feature is used. An app partner can check integration compatibility with the current theme. Those are proposed research steps, not purchasing opportunities demonstrated by this study.

Keep the prior/current identity, observation dates, visible feature and unresolved question in a note. Rule out renaming or detector changes, ask about the actual workflow, and verify the relevant contact before proposing a project. Traffic, theme and qualification filters help narrow candidate accounts; they cannot establish a merchant is mid-migration or ready to buy.

For service-specific checks, see the redesign guide and app combinations. A current recommendation should use current feature documentation and measured requirements, rather than select a theme because its April label count rose.

FAQ

Which theme gained the most in this study?

Horizon had the largest named-product absolute gain in the selected result, +84. Tinker had a higher percentage increase. Broad Custom classification and literal Custom names are separate measures.

Does a 120-day panel mean every store was tracked for 120 days?

No. It requires at least two qualifying snapshots inside that lookback. Exact elapsed spans and overlap were not retained.

Does a theme change mean a merchant will buy adjacent apps?

The study did not measure that. Verify the current identity, task, relevant contact and merchant interest before making an offer.

Does this describe current theme performance?

No. It reports April 17 historical theme-label changes, with no performance, conversion or buyer-budget measurements.