E-commerce
October 30, 2026
Wondering how custom views and filters in Google Analytics can help you dissect your visitors' behavior?
In just a few minutes, you will know how to create targeted views, apply reliable filters, and gain actionable insights to boost your sales.
Don't let raw data fool you: without structure, you risk missing key signals and making erroneous decisions.
So, how do you leverage custom views and filters in Google Analytics? Here is the agenda:
Why create multiple views for the same e‑commerce site?
How to design a filter that eliminates redundant URLs?
What impact does the order of filter application have on your reports?
How to test a filter before deploying it to production?
How Qstomy uses filtered data to improve your purchase funnel?
Let's go.
Throughout this article, we will also detail best practices for data governance, common pitfalls to avoid, and advanced tips for combining multiple filters without losing granularity. You will leave with a clear roadmap ready to be applied immediately to your Google Analytics account.
Summary
Why create multiple views for the same e‑commerce site?
Each view in Google Analytics acts as a dedicated perspective on your traffic. By segmenting data—for example, one view for organic traffic and another for paid campaigns—you obtain clearer and more actionable reports. For e‑commerce, this means being able to track separately the behavior of visitors from your newsletters, Facebook ads, or internal search. This granularity helps you identify the levers that generate the best return on investment and adjust your strategy accordingly.
In practice, create a "Transactions Only" view that excludes all non‑transactional traffic; this will allow you to calculate KPIs such as average revenue per visitor without being clouded by purely informational visits. Similarly, a "Mobile" view will give you visibility into your site's performance on smartphones, which is indispensable in the mobile‑first era. By combining these views, you build a comprehensive dashboard that meets the requirements of your marketing, product, and finance teams.

Convert over 2,000 customers on average per month with Qstomy.
The world’s 1st Shopify AI dedicated to customer conversion



Empowering 200+ e-commerce merchants
How to design a filter that eliminates redundant URLs?
On many stores, the same product page appears under multiple URLs, such as /index.php/produit and /produit. A search‑and‑replace filter can remove the segment /index.php so that Google Analytics groups both visits under a single URL. In the interface, select "Custom Filter", type "Search and Replace", field "Request URI", search for /index.php and replace it with an empty string. Thus, your reports display the actual number of views per product, without double counting that would skew your assortment decisions.
To go further, you can extend the same principle to unnecessary tracking parameters (e.g., ?utm_source=mail) by removing or standardizing them. Also consider normalizing URL casing so that /Produit and /produit are treated as identical. These adjustments reduce analytical noise and offer you a more reliable view of the actual popularity of each SKU.
What impact does the order of filter application have on your reports?
Google Analytics processes filters in the order they are listed. If you first apply a filter that keeps only US traffic, and then a second that keeps only Canadian traffic, the result will be null. Each filter only sees the output of the previous one. To avoid this trap, combine criteria within a single filter or organize the order logically: group global inclusion filters first, followed by specific exclusions.
An advanced tip is to place cleaning filters (removing parameters, normalizing URLs) before inclusion filters; this ensures data is homogenized before being filtered. Test the order by creating a test view and comparing metrics with the raw view; you will quickly see the impact of each permutation on conversion rates and session counts.
How to test a filter before deploying it to production?
Before applying a filter to your main view, create a test view. Add the filter and review reports for 24 to 48 hours. Ensure the expected traffic appears and that key data is not lost. If the filter contains an error, you can still remove it from the test view without impacting your historical data. This preventive step avoids irregularities that, once applied, are irreversible.
During the test phase, use real-time reports to validate that filtered visits are correctly excluded or included as expected. Also compare quality metrics (bounce rate, average session duration) between the test view and the raw view to detect any biases introduced by the filter. Once validation is conclusive, replicate the filter on the production view and document the change in your data governance log.
What filter ideas are most useful for e‑commerce?
Here are some filters that add real value to your online store:
Campagnes email uniquement : include the UTM parameter from your newsletters to measure their real impact.
Returning visitors : filter sessions with an existing cookie to analyze customer loyalty.
Traffic from banner ads : exclude IP addresses from ad networks to keep only qualified traffic.
Social referrers : capture visits from your Facebook, Instagram, or TikTok pages to optimize your content.
These filters allow you to focus your analysis on the levers that truly drive conversions.
Additionally, you can create a filter for 'Transactions excluding shipping fees' that retains only orders with a net amount exceeding a defined threshold, to isolate high-value sales. Another effective filter involves excluding bot and crawler traffic by relying on user-agent identifiers, which improves the accuracy of real human behavior metrics.
Why keep a raw view with unfiltered data?
Filters can create invisible errors: a poorly constructed regular expression may exclude legitimate visits. By maintaining an unfiltered view, you always have a source of truth to verify the results of your filtered views. If you suspect an anomaly, compare the figures with the raw view. This practice is especially important when feeding your attribution models or CRM reports, as explained in e-commerce CRM and customer support.
Furthermore, the raw view serves as a foundation for ad-hoc analyses, such as identifying new customer segments or calculating custom metrics that were not anticipated during the initial filter setup. Remember to regularly back up an export of this view to ensure secure archiving, especially before any major changes to the filtering schema.
Views do not roll up historical data: what to do?
A filter applies only from the moment the view is created. To avoid gaps, plan your views from the launch of your store. If you need to segment retroactively, export the raw data via the API or a CSV, then recreate the filters in a new view. This way, you maintain analytical continuity and do not lose valuable information about past performance.
It is also possible to use the segment import features in Google Analytics 4 to apply filtering criteria retroactively to historical data. This approach allows you to reconstruct before-and-after comparative reports and identify seasonal trends without having to start from scratch.
How to create a new reporting view in Google Analytics?
Go to the menu Admin, select the desired account and property, then click the dropdown menu Vue and select Create a viewDon't give it a evocative name (e.g., 'Organic Traffic') and set the corresponding time zone for your store. Validate and the view will start collecting data from that moment. Think about adding an exclusion filter for internal traffic right from the start, then create the specific filters you need. To go further, consult our guide on creating a payment gateway, which shows how to align your transactional data with your GA views.
After creation, configure your goals (conversion, sign-up, etc.) and associate relevant segments. A best practice is to enable custom event tracking from the first view to avoid losing crucial information during future site updates.
Predefined filters or custom filters: when to choose which?
Google Analytics offers predefined filters for common needs, such as excluding internal traffic or filtering IP addresses. They are simple to set up and avoid syntax errors. However, custom filters offer total flexibility: you can combine search-replacement, include/exclude URL parameters, or even create rules based on event fields. For example, to group URLs /index.php/produit and /produit, only a custom filter will work. Use predefined filters for standard tasks and reserve custom filters for your store's specific needs.
When choosing a custom filter, test it first on a test view and validate the result with the raw view. Document each filter, its purpose, and its syntax in a shared repository so your entire team can understand and reproduce the configuration if needed.
What pitfalls to avoid when filtering GA data?
Over-filtering is the most common danger: by excluding too much traffic, you reduce your database and skew your conversion ratios. Always ensure each filter has a clearly defined objective. Additionally, do not forget the order of application: a too-restrictive inclusion filter placed before an exclusion filter can eliminate visits you wanted to keep. Finally, test each new filter in a backup view before applying it to your main view. For SEO compliance, follow our recommendations in l’optimisation Google.
Another pitfall is reusing the same filter across multiple views without adaptation; contexts differ (mobile vs. desktop, organic vs. paid) and a too-generic filter can mask relevant signals. Adopt a modular approach where each filter has a precise scope and is documented in your data governance plan.
How Qstomy uses filtered data to optimize your purchase funnel?
Qstomy relies on filtered Google Analytics views to refine every step of the customer journey. By isolating traffic from email campaigns, we identify the most popular products and propose upsell recommendations directly in the cart. Data from returning visitors allows us to personalize follow-up messages and adjust cross-sell options. Moreover, thanks to a view without internal traffic, our stock forecasting algorithms are more reliable, reducing stockouts and improving package tracking. Finally, customer service benefits from a clear overview of recurring issues through the 'visitors who abandoned the checkout funnel' segment. Discover how to choose the best e-commerce platform on notre comparatif plateforme.
We also enrich our dashboards with post-purchase engagement metrics (product reviews, social shares) using filters that isolate users who made a purchase and then interacted with content. This 360° view allows us to adjust re-engagement strategies and measurably increase customer lifetime value (CLV).
What checklist to follow before deploying a new filter and quick FAQ?
Checklist before deployment
Create a separate test view.
Clearly define the filter's objective.
Verify syntax (regex, parameters).
Compare results with the raw view.
Plan the order of application with other filters.
FAQ express
Can a deleted filter be restored? No, it must be recreated; that is why the test view is crucial.
Do filters affect real-time reports? Yes, filtered views display real-time data based on the applied filter.
Should I create a view per campaign? Not necessarily; group similar campaigns in the same view to keep management simple.
To go further on organic traffic, consult our e-commerce SEO guide ou our acquisition strategies. Bonne configuration !
To go further: Training an e-commerce chatbot with Shopify: using the right data without creating bad responses - Qstomy.
Finally, remember to document each filter in your internal registry, noting the creation date, the responsible person, inclusion/exclusion criteria, and tracking indicators. This traceability facilitates audits and ensures compliance with data protection requirements.

Enzo
October 30, 2026


