Shopify Store Benchmarks [1M-Store Study]

TL;DR

• The October 25 refresh compares 1,053,913 store-domain records with successful full scans from September 25 through October 24. It is a recent coverage cohort, not a merchant census.
• The cohort's median is 1 retained app name and 2 retained pixel names; the means are 2.19 and 3.44. These do not prescribe an ideal stack.
• 97.00% of the cohort is in the estimated under-50K traffic tier. The 1M–5M group contains only 26 records.
• March and April tables remain dated historical context. Different samples, age rules and detection definitions prevent a growth comparison.

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 useful benchmark gives you a comparable cohort and a question to investigate. It does not prescribe a number of apps to buy or convert estimated traffic into an agency budget. The October cohort and retained March/April studies use different selections and counting rules. Keep their denominators separate.

October 25 refresh: recent traffic and stack benchmarks

We extracted one consistent cohort on October 25, 2026: 1,053,913 store-domain records, each with a successful full snapshot from September 25 through October 24 (UTC observation dates). We selected the latest qualifying snapshot per record, required recorded app and pixel lists, and excluded 3,048 records currently marked confirmed dead. Scans with no detected apps or pixels remain in the denominator. This is our recent scan coverage, not a representative census of live Shopify merchants.

Traffic tiers come from the same snapshots. The scan timestamp does not prove that a store still trades, that every signature is accurate, or that a detected app is enabled. Detector versions were unrecorded in this cohort, so we cannot standardize detection changes across the older and newer studies. The retained query and aggregate output record the counting rules and observation window.

For this table, we count distinct normalized app and pixel names within each selected snapshot. App totals exclude payment-provider identities from the current signature catalogs and four withdrawn generic detections: Cart Drawer, Shopify, Cart Functionality and SummerCart. Pixel totals exclude five identities whose signatures do not support an external adoption claim: Yahoo Advertising, Triton Digital, Retention.com, Foursquare Attribution and Kroger Precision Marketing. These are explicit research counting rules, not paid-subscription counts or a claim to have detected every tool.

Across all 1,053,913 records, the means are 2.19 apps and 3.44 pixels, and the medians are 1 and 2. 331,830 records have no retained app name under this definition. Zero names does not prove zero apps or missing functionality.

The cohort contains no record in a recorded 5M+ tier. That is a limit of this scan window, not a claim that those stores do not exist. With 97.00% of records in the lowest tier and only 26 in the highest observed tier, use a relevant peer group rather than the overall mean. A merchant with 50K–200K estimated traffic can compare with that group's median of 5 app names, then inspect which customer tasks those tools serve. Buying enough apps to match the median has no demonstrated benefit.

We have not followed the same merchants from March to October, and detector versions are unavailable for the October cohort. The larger research sample and different counting rules therefore do not show that merchants gained apps, lost pixels or changed their business performance.

The March Traffic-Tier Benchmark

StoreInspect retained this extraction on March 1, 2026. The five traffic groups sum to 191,980 records. The query grouped stored store fields by recorded estimated traffic tier and calculated means from available app and pixel counts. Query averages skip missing values; the retained summary does not preserve how many records had a value for each mean, so the group size is not a verified denominator for every average.

This is a database sample, not a representative census of active Shopify merchants. No active-store rule or maximum observation age was applied in the retained query. These are stored historical fields, not a claim that every storefront was visited on March 1. Detection rules and app counting can change; the current product excludes payment-provider entries from app totals, whereas these historical counts used the earlier rules.

83.3% of these records were in the under-50K group. That is a distribution of stored traffic estimates, not a failure rate. A low-traffic store may be new, specialized, wholesale or reliant on channels this estimate does not represent.

Mean apps increase across these groups, but mean pixels do not increase at every step: 8.4 in the 200K–1M group versus 8.3 in the 37-record 1M–5M group. These are different stores observed cross-sectionally. The data does not show the same merchants adding apps as they grow, whether apps caused growth, or the appropriate app count for a particular business.

Use the two larger groups for a broad historical comparison, with the counting and coverage limits in mind. Do not choose an enterprise tool simply because its potential users fall in one of the tiny upper groups.

Pixel Signals: Presence Is Not Delivered Measurement

The retained March query searched any stored snapshot for the named pixel string. It did not restrict the search to each store's latest snapshot. A match may therefore be historical even at the extraction date. The denominator was the stored traffic-tier group, rather than a set of manually confirmed active installations.

These rows describe detected historical signatures. They do not measure currently delivered events, consent behavior, accurate attribution or active campaigns. A Google Ads signature is not evidence that the merchant spends more than one with a TikTok signature. The table also cannot establish overlaps between two tools from their separate shares.

A Universal Analytics remnant needs especially careful interpretation. Google's transition documentation explains that standard UA processing ended in July 2023 and the 360 extension ended in July 2024. Legacy code does not demonstrate active UA collection or a failed GA4 migration. Investigate current event delivery with authorized access rather than pitching a migration from a leftover script.

For the operational checks, see our pixel detection guide. Shopify's consent-dependent and sandboxed pixel behavior means a top-level script check alone cannot verify every implementation.

Category Benchmarks From a Separate April Cohort

The April 30, 2026 CRO extraction found 202,409 stores in recorded traffic tiers of 50K or above, from 560,928 records with eligible latest snapshots. It then selected 176,305 with an ad-related signal: a positive stored Meta ad count or a Meta/Facebook, Google Ads, TikTok or Pinterest pixel pattern. That signal does not prove current campaigns or spend. It selected each store's latest available snapshot without a maximum age. The category table below uses recorded StoreInspect classifications; it does not represent all stores in those industries.

The workflow intersection used category flags for email marketing, reviews and upsell. It did not measure whether the merchant sends campaigns, collects reviews or already provides native recommendations. The five displayed categories are selected examples, not an exhaustive niche ranking.

Beauty's mean of 10.4 exceeds the other four means shown here. That does not establish the most advanced niche, higher margins or the reasons for the difference. App counts depend on historical signatures and counting rules; they are not paid subscriptions or engineering quality. Do not compare April means directly with March means as a trend: the selections and extraction dates differ.

The 1,118 Home records in the email/reviews/no-upsell intersection are an example of a practical shortlist. Inspect product compatibility, room sets or spare parts, and check existing recommendations before proposing a service. For the 984 Beauty records, inspect shade, routine and variant requirements. Those specific checks are more useful than a generic claim that a store is below average.

Themes, Plus and Scores Need Their Own Validation

A free theme does not prove a limited budget, and a custom/unknown label does not prove a bespoke build. Some features may come from the theme, native Shopify tools, custom code or services that do not expose a public signature. Evaluate the actual product selection, navigation, accessibility and performance requirements.

A stored Shopify Plus flag is also a detection signal. Confirm a current plan when a recommendation depends on Plus-specific checkout functions. Subscription selling does not by itself require Plus. Platform charges do not establish spare project budget, and a score that includes Plus cannot provide independent evidence of Plus causing success.

Lead fit scores combine signals for prioritization. They do not establish revenue, profit, willingness to buy, a contact's purchasing authority or consent to outreach. Use them to order research only after defining a service-specific qualification rule. Our store example guide shows why actual customer tasks and provider evidence matter more than a universal rank.

Build a Service-Fit Pool Before Calling It a Market

“No covered app detected” is a candidate condition, not an absent workflow. The retained studies do not provide a qualified total addressable market. Inspect the current store and ask:

1. Is a native, theme, custom or backend solution already handling the task?
2. Does the merchant have a demonstrable problem, such as confusing variants, delayed responses or an incomplete reporting workflow?
3. Can the proposed service address that problem within operational, margin and data constraints?
4. Is there a current relevant contact, owner and confirmed budget?
5. Has the account opted out, become an existing customer, or entered another active conversation?

Track how many candidates survive each step. A contact record is not a reachable decision-maker, and a stored verified status is not a guarantee of current deliverability. Do not multiply a historical category count by an unrelated contact percentage.

Revenue and Product Prices Are Different Benchmarks

No retained independent sales validation establishes actual revenue for these cohorts. A model based on traffic, catalog and pricing inputs is an estimate; it cannot justify a claim that a store can afford an agency retainer. Discuss actual budget with the merchant.

For a revenue benchmark, use a clearly specified period and currency, distinguish gross sales from refunds and profit, and report the source and uncertainty. Product-price means also require normalized currencies and units, coverage and robust summaries. A large raw mean could reflect data problems as well as expensive products. Similar means and medians do not establish a normal distribution.

Our revenue analysis explains those limits and provides a separately dated historical view. These app and pixel tables do not add sales evidence.

Use the Benchmarks for an Inspection

For a store owner, choose one customer or operational task and inspect whether the current stack serves it. For an agency, use a specific intersection to prioritize research, then validate the problem before drafting a pitch. For an app developer, examine integration and workflow requirements rather than assuming all non-detections are buyers.

The StoreInspect dashboard can help discover candidates. Record the filter, export date and observation dates so a later comparison is reproducible. Current filters and exports depend on the plan; recheck the product's entitlements before designing an export workflow.

Frequently Asked Questions (FAQ)

How many apps should a Shopify store have?

There is no measured optimum here. The October cohort has a median of 1 retained app name and a mean of 2.19; its traffic-tier medians differ. March means use separate historical counting rules. Choose apps for a verified need and check redundancy, performance and total cost.

Does a higher-traffic store need more pixels?

The larger March groups had more historical pixel matches on average. That does not prescribe a pixel quota. Install measurement needed for your campaigns and reporting, respect consent, and test actual events.

Can missing-app filters identify agency clients?

They can identify stores to investigate. Native or hidden tools may already provide the capability; budget, role, workflow need and current contact details remain unconfirmed.

Are these quarterly updated benchmarks?

The latest table was extracted October 25, 2026, using successful full scans from September 25 through October 24. Historical tables retain their own dates. No automatic or guaranteed quarterly refresh is implied. For a fresh comparison, use a consistent cohort and counting definition rather than comparing unrelated database snapshots.