How Third-Party Apps Affect Shopify Speed: Data from 1,533 Stores

Many Shopify apps and third-party services add JavaScript or other external requests to your storefront. That includes ad pixels, analytics tags, chat widgets, review apps, and BNPL providers. These services often add requests and main-thread work during page load or shortly after, competing with core storefront resources for browser time.

We measured how strongly this third-party load is associated with mobile performance across 1,533 Shopify stores in 9 industries (from our Shopify Speed Benchmark). The metric here is not "number of Shopify apps installed" but the number of distinct third-party services detected on each store's homepage, including app scripts, tracking tags, embeds, analytics, CDNs, and other external resources.

In our dataset, stores with fewer detected services scored higher on mobile. The most common range is 21-30 third-party services, and stores in that group score 44 on mobile Lighthouse. Stores with 5 or fewer score 74. That 30-point gap suggests a meaningful difference in storefront weight and execution cost, even if Lighthouse score is not a direct business metric. A Deloitte and Google study found that a 0.1-second improvement in mobile load time increased retail conversions by 8.4%.

This article breaks down which types of services stores load, which ones block the browser the most, and how to audit your own store.

How Many Third-Party Services Do Shopify Stores Load?

We identified every third-party service loaded on each store's homepage using Lighthouse entity detection, then grouped stores by service count. The Pearson correlation between service count and mobile performance score is r = -0.66, a strong negative relationship.

The drop is steepest between 0-5 and 11-15 services. Going from 5 to 15 services cuts your score by 15 points. After 30 services, each additional one has a smaller effect because the score is already so low.

Most stores (460 of 1,533) fall in the 21-30 range. Only 42 stores run with 5 or fewer services.

What Types of Services Stores Load

Across the 1,533 stores in our dataset, we categorized 37,274 individual third-party entity instances. The chart below shows the share of each category. Note that these are shares of all detected entity instances, not shares of stores.

Ad and analytics scripts together make up a third of all third-party load. This is worth noting because ad pixels and analytics tags are often the first things added to a new store and the last things anyone thinks to audit.

CDN services rank second, but most of these (Shopify's own CDN, Cloudflare) actually help performance rather than hurt it. The more meaningful performance costs in this dataset come from the other categories.

Which Service Categories Affect Performance Most

Not all categories of third-party services have the same relationship with performance. In our benchmark data, we correlated each category's count per store with the store's mobile Lighthouse score across all 1,533 stores.

Analytics services show the strongest negative association with performance (r = -0.55). This is partly because stores with more analytics tools tend to be stores that also load more of everything else. But analytics scripts are often loaded early in the page lifecycle to capture page view data, which can add to execution cost.

Ad pixels come in at r = -0.41. Stores running Facebook Pixel, Google Ads, TikTok Pixel, and Pinterest Tag simultaneously are loading four separate tracking libraries that each need main-thread time.

Checkout and review services show the weakest correlations (r = -0.12 and r = -0.20). These categories tend to load fewer scripts per store, and checkout-related services like Affirm and Klarna may have less homepage impact than analytics or marketing tools, especially when they load later in the shopping journey.

One note on interpretation: these are correlations, not isolated causal measurements. A store with 5 analytics services likely also has more ad pixels and marketing tools. The category-level numbers show which types of services tend to co-occur with lower scores, not that removing one analytics tag will improve your score by a specific amount.

Which Services Block the Browser the Most

Not all third-party services are equal. Some load quickly and stay out of the way. Others take over the browser's main thread for a full second or more.

We measured main thread blocking time from Lighthouse's third-party audit across all 1,533 stores. The numbers below show how long each service blocks the browser during page load (p25 to p75 range):

Signifyd stands out. On stores where it appears, it contributes a median of 1.2 seconds of main-thread work. That level of blocking can delay interactivity during page load and make the page feel less responsive.

Google Tag Manager is the most widespread service at 90% adoption, and it blocks the main thread for a median of 486ms. That half-second matters because GTM often loads additional scripts (Facebook Pixel, Google Analytics, Klaviyo) that add their own blocking time on top.

Rebuy and Affirm are comparatively lightweight. Both stay well below the median blocking time of the heavier services in this list.

The relationship between service count and main-thread blocking is consistent across our dataset. The Pearson correlation between entity count and Total Blocking Time is r = 0.43, a moderate positive relationship. More services means more JavaScript competing for the main thread, which tends to increase delays before the page becomes fully interactive.

Google Tag Manager: The 90% Problem

GTM deserves its own section because 90% of stores in our dataset use it (1,383 of 1,533).

GTM itself is not the problem. The problem is what stores load through GTM. Many Shopify stores use GTM to fire Facebook Pixel, Google Analytics 4, TikTok Pixel, and sometimes Klaviyo or Hotjar. Each of these scripts loads and executes through GTM's container, stacking blocking time.

Google recommends GTM as a way to consolidate tags and load them asynchronously. That recommendation is sound. But the benefit disappears when merchants keep adding tags without removing old ones.

The audit step is straightforward, and we cover it in more detail in our site speed best practices guide: open your GTM container, count the active tags, and ask whether each one is still providing data you actually use. If you stopped checking your Facebook Ads dashboard six months ago, the pixel is still running on every page load.

Third-Party Density Varies by Industry

Some industries simply load more services than others. The chart below shows the median number of third-party entities detected on each store's homepage, not the full tech stack behind the scenes. Beauty stores load roughly twice as many as automotive stores:

Beauty and personal care stores tend to use more review apps, loyalty programs, quiz tools, and marketing automation than other industries. Each one adds its own scripts. For a full breakdown of which services are most concentrated in beauty and how they affect performance, see our beauty speed benchmark.

Automotive parts stores are leaner on average, which may reflect a more functional storefront stack and fewer engagement-layer tools. Fashion & Apparel stores sit in between, with return-prevention tools like Kiwi Sizing (74% of all benchmark installs in Fashion) and Route (572ms blocking) adding a performance layer that other industries don't carry.

This matters because a beauty store owner comparing their score to an automotive store is comparing two very different baseline script loads. For a deeper look at how industries compare on performance and Core Web Vitals, see our site speed optimization guide.

How to Audit Your Shopify Store's Third-Party Script Load

You can measure your own store's third-party impact in about 10 minutes.

Step 1: Run a Lighthouse audit

Open Chrome DevTools (F12), go to the Lighthouse tab, and run a Performance audit on your homepage. Look for the "Reduce the impact of third-party code" section. It lists every external service, how much transfer size it adds, and how long it blocks the main thread.

Alternatively, run your store through EcomHint's speed test to get the same data with ecommerce-specific context and a benchmark comparison.

Step 2: List every service and its cost

Create a simple spreadsheet with three columns: Service Name, Main Thread Blocking Time, and Whether You Still Need It. Anything blocking the main thread for over 250ms is worth investigating.

Step 3: Decide what stays

For each service, ask:

• Is this generating revenue? (Affirm for BNPL, Rebuy for upsells) If yes, keep it. The revenue likely outweighs the speed cost.
• Is this providing data you check weekly? If you haven't opened the dashboard in months, remove the script.
• Can this load later? Chat widgets, review widgets, and social proof popups don't need to load before the page is visible. Use facade patterns to defer them until user interaction.

Step 4: Clean up your GTM container

Open Google Tag Manager and review every active tag. Remove tags for services you've uninstalled. Consolidate duplicate tracking (many stores fire both a Shopify pixel and a manually added pixel for the same platform).

Step 5: Check for leftover app code

When you uninstall a Shopify app, it sometimes leaves its script behind in your theme code. Go to Online Store > Themes > Edit Code and search for <script tags or <link tags referencing services you no longer use.

Third-party script load is one part of a broader Shopify store audit that also covers SEO, mobile UX, trust, CRO, and prioritization in a 6-step process.

The theme-code and GTM steps can get fiddly, and it is the kind of work that WebDesk Solution's Shopify speed optimization service takes on for merchants who would rather not do it in-house.

The Revenue Perspective

Removing third-party services is not always the right answer. Some services earn more than they cost in speed.

Affirm (buy now, pay later) blocks the main thread for a median of 183ms. That is barely noticeable. But stores offering BNPL options often report higher average order values. The speed cost is typically worth it.

Rebuy (upsell engine) blocks for 222ms and is used by 11% of stores in our dataset. If it generates even one additional product per order, the conversion math works.

The services that hurt are the ones providing no measurable return. A Facebook Pixel for a campaign you stopped running. A Hotjar recording you never watch. A chat widget nobody uses. Each one costs speed for zero benefit. Product pages are especially sensitive to this because they already load more JavaScript than homepages. Our product page speed analysis shows PDPs score lower in 8 of 9 industries.

What the Research Says About Speed and Revenue

The connection between page speed and ecommerce revenue is well-documented:

• A Deloitte and Google study of 37 sites and 30 million sessions found that a 0.1-second improvement in mobile load time increased retail conversions by 8.4%.
• Vodafone ran an A/B test where a 31% improvement in their main content loading time led to 8% more sales, with no other changes to the page.
• eBay found that every 100ms improvement in search page load time increased Add to Cart actions by 0.5%.

These case studies support the broader pattern that reducing real speed bottlenecks can improve ecommerce outcomes, though the size of the uplift depends on the page and the test design.

Our benchmark data shows a 30-point gap between stores with 0-5 services (median score 74) and stores with 21-30 services (median score 44). Lighthouse scores are not a direct proxy for revenue, but external case studies suggest that meaningful speed improvements can translate into better conversion outcomes when they improve real user experience.

FAQ

How many Shopify apps is too many?

There is no fixed number. In our dataset, 460 of 1,533 stores run 21-30 third-party services and score a median of 44 on mobile. Stores with fewer services score substantially better: 74 in the 0-5 range and 68 in the 6-10 range. The question is not how many apps you have, but whether each one provides enough value to justify its speed cost.

Do Shopify apps slow down my store even if I disable them?

Sometimes. Uninstalling an app from Shopify removes the app itself, but some apps leave their script tags behind in your theme code. Check your theme's code editor for any <script> tags referencing services you no longer use.

Which third-party services have the biggest impact on speed?

In our data, fraud prevention tools like Signifyd (median 1,209ms blocking time) and tag managers like GTM (486ms) have the largest main thread blocking times. Marketing popups like Privy (453ms) and shipping protection like Route (412ms) also add noticeable delay. BNPL services like Affirm (183ms) and upsell engines like Rebuy (222ms) are relatively lightweight.

Should I remove Google Tag Manager to improve speed?

Not necessarily. GTM itself adds around 486ms of blocking time. But removing GTM means you need another way to manage your tracking tags, and adding them individually can be worse. The better approach is to audit what runs inside GTM and remove tags you no longer need.

Does the number of products in my catalog affect speed?

Not meaningfully. Across the 1,533 stores in our dataset, the correlation between product count and mobile performance score is near zero (r = -0.03 for homepage, r = -0.08 for collection pages). Stores with 50 products scored about the same as stores with 5,000. Shopify paginates collection pages, so catalog size does not change how much content renders on any given page load. What does affect speed is the number of third-party services, theme weight, and media on the page itself.

Which types of third-party services hurt speed the most?

In our data, analytics services show the strongest negative association with mobile performance (r = -0.55), followed by utility scripts (r = -0.46) and marketing tools (r = -0.45). Checkout services like payment providers show the weakest association (r = -0.12). These are correlations across 1,533 stores, so they reflect broader patterns rather than isolated per-service impact.

Methodology

This analysis uses data from the EcomHint Shopify Speed Benchmark. The dataset includes 1,533 Shopify stores across 9 industries, each with CrUX field data and at least 3 successful Lighthouse runs on both mobile and desktop.

Third-party services were identified using Lighthouse's entity detection system, which recognizes services by their script URLs and network patterns. Entity counts reflect unique services per store, not individual script files.

Blocking time was measured from the mainThreadTime field in Lighthouse's third-parties-insight audit, using the median of 3 mobile homepage runs per store. Values are reported as p25-p75 ranges across all stores where each service was detected.

The full dataset is available for download in the benchmark article.