E-commerce
October 28, 2026
Wondering why your Google Analytics figures don't match your store's reality? It's normal: the tool classifies and attributes traffic according to internal rules that produce counter-intuitive data. As a result, you may make marketing decisions on flawed foundations.
Google Analytics is powerful, but this power comes from its flexibility. And this flexibility creates blind spots: direct traffic absorbs visits that aren't direct, the same visitors are counted twice, and third-party domains slip into your referrers. Before optimizing anything, you need to know where to look.
So, how to understand confusing Google Analytics data? Here's what's covered:
How does Google Analytics actually attribute your visitors?
Why is direct traffic overestimated?
Faut-il raisonner en visiteurs ou en sessions ?
Why does your payment show up as referral traffic?
What checklist to apply before concluding an analysis?
Let's go.
Summary
How does Google Analytics actually attribute your visitors?
Google Analytics sorts your traffic into simple channels: organic search, referral, email, social media. It's practical on the surface, but misleading in detail. Because this ranking isn't based on what the person actually did, but on two technical variables: 'source' and 'medium'.
If a session's medium is 'organic', the session is categorized as organic search, even if the user arrived via a detour. The 'Direct' channel, meanwhile, groups everything without a defined medium: a manually typed URL, a saved bookmark.
The problem with automatic rules
These definitions are documented by Google, but they remain opaque to a non-technical merchant. You see a clean report and believe it's the truth. In reality, it's an interpretation. Keep this principle in mind: a channel in Google Analytics is a hypothesis, not proof.

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
Why is direct traffic overestimated or underestimated?
The Direct channel has the lowest priority in attribution. Concretely: if a visitor first arrives via organic search, email, or a campaign, then returns later by typing your address directly, this second visit is still credited to the original channel.
The impact on your figures
Direct consequence: organic search, email, and other indirect channels appear more effective than they truly are. Conversely, direct traffic is undervalued, even though it often reflects established brand awareness.
The reverse also exists: a visitor who first arrives directly and then via an indirect channel will be correctly attributed to the indirect channel. You therefore lose sight of the spontaneous brand awareness that brings your customers back, while overvaluing channels that simply reached a visitor who was already converted.
Faut-il raisonner en visiteurs ou en sessions ?
Google defines a session as a group of interactions on your site within a given time window. A visitor, however, is a person. This distinction may seem trivial, yet it changes all your reports.
Consider a concrete case from classic analyses: a segment based on sessions that started on the homepage shows 291 sessions in September. The same report, now based not on sessions but on visitors who entered via the homepage, shows 394 sessions for that same month.
Why such a discrepancy?
The 'session' segment only retains sessions that began on the homepage. The 'visitor' segment retains all people who passed through the homepage at least once during the month, including subsequent sessions that started elsewhere.
These visitors returned at other times and generated multiple sessions. Your report then counts landing pages that have nothing to do with the homepage. Result: two reports, two truths, and a decision that can go in the wrong direction.
Why does your payment appear as referral traffic?
An e-commerce site often relies on third-party domains during checkout. With a solution like PayPal, the buyer leaves your store to complete payment and returns once the transaction is finished. This break disrupts the tracking chain.
Result: upon the payment provider's return, the visit is classified by default as referral traffic, as if an external site had sent the customer. Your referral report then fills up with domains that are not true partners.
The fix
The solution is simple once identified: add these domains to an exclusion list in your analytics tool. Without this list, you believe you have a new acquisition source when it is merely a mandatory step through the checkout. Systematically verify all third-party domains involved in your order funnel.
Why do your subdomains skew referrals?
Another common source of confusion: your own domain appears as a major referrer. The blog, help center, or a secondary version of your address present themselves as if they are sending traffic to your main site.
If your store operates across multiple subdomains, such as a blog, an info section, and the main site, these subdomains can generate self-references. The visitor navigates internally, but the tool interprets this as a new external visitor.
The real cause
Most often, the tracking code is misplaced or duplicated on certain pages. You must regularly audit the tracking code across all subdomains and pages. A quarterly check prevents building months of reports on phantom referrers that are, in reality, your own site.
What is referral spam and how to spot it?
A persistent issue with Google Analytics: the arrival of fake referrers that pollute your metrics. Spammers detect your tracking ID and then use automated bots to trigger this code remotely.
The tool records a session even though no real visit occurred. The bot merely accessed the tracking code. These intrusions inflate your referral traffic and severely disrupt the tool's usage.
How to protect yourself
The countermeasure involves creating filters to exclude these parasitic domains and maintaining up-to-date segments to isolate only legitimate traffic. Without this hygiene, your conversion rates and cost per acquisition by channel become unusable. Make it a habit to review your referrer list monthly, when the data is fresh.
Why do your conversion data seem inconsistent?
Combine these biases and you get the classic scenario: one channel appearing highly performant, another underestimated, and a total conversion count that doesn't match your actual sales. This isn't a calculation error; it's a sum of small drifts.
The undervalued direct channel hides your brand awareness. The overvalued organic traffic pushes you to invest where the effort isn't necessarily the most profitable. Referrer spam adds noise you risk mistaking for real opportunities.
The right reaction
Never correct an isolated channel without checking the whole picture. Treat this data as material to clean, not as a verdict. To go further on reading usage by role, consult our article on how data analytics serves e-commerce.
How to link your analytics data to your customer relationship?
A corrected traffic data point has value only if it aligns with your other store signals. Knowing which channel truly brings a buyer changes how you handle their history, questions, and follow-up.
A customer arriving via email doesn't have the same journey as one arriving via organic search. If your sources are skewed, your team responds off-target. This is where cross-referencing with your customer relationship tools becomes useful.
A consistency issue
To structure this reading, rely on the right support data, as in our guide on the e-commerce CRM and customer support. A customer service team fed with clean data transforms an anonymous visitor into a loyal customer, and your figures become a decision-making tool again.
Why does your platform choice impact data quality?
Not all platforms behave the same way regarding tracking, third-party payments, and subdomains. A store built on a lightweight foundation won't expose the same pitfalls as a heavier installation.
If your payment funnel relies on external services, prepare for parasitic referrers. If you multiply subdomains, prepare for self-references. These constraints must factor into your tool selection, not just your marketing strategy.
Choosing with full knowledge
Compare options based on your actual ability to audit tracking, for example via our article on which platform to choose to create an e-commerce site. A suitable platform reduces analytical noise even before the first correction.
How do your data also serve customer service?
Understanding an acquisition source changes how you welcome a customer. You know if they are discovering you, returning, or coming from a partner. This information directly feeds your responses, priorities, and tone.
This also applies to assistance tools. An assistant fed with clean data avoids generating wrong answers, as explained in our guide for training an e-commerce chatbot with Shopify.
The right tool at the right time
Depending on the need, a FAQ, internal search, or conversational assistant won't play the same role. Our comparison on FAQ, recherche ou chatbot IA helps you decide. And don't forget the human aspect: the social media for e-commerce customer service extend this relational logic.
How does Qstomy help interpret your data to convert?
Qstomy is an AI agent for Shopify that guides the buyer from discovery through post-purchase. Where Google Analytics describes traffic upstream, Qstomy acts on what happens once the visitor is in your store: recommendations, cart building, and cross-selling.
The agent also handles order tracking, account access, and policy questions, limiting unnecessary back-and-forth with your support. Over 100 merchants already use it daily.
Two complementary roles
Google Analytics tells you where the visit comes from. Qstomy works on conversion, cart, and after-sales service. A well-designed payment funnel and a reliable gateway complete the setup, as detailed in our article on creating an e-commerce payment gateway.
The idea isn't to oppose the two: your data cleans the vision, the agent converts the visit.
What checklist to apply before concluding an analysis?
Before drawing a conclusion from a report, run these simple checks. They prevent building a decision on skewed data.
Exclure the third-party domains used in the payment of your referrer list.
Auditer tracking codes across all subdomains and pages.
Filtrer spammy referrers and keep your segments up to date.
Choose consciously between session-based and visitor-based reading.
Interpret treat the direct channel as a signal of brand awareness, not as a void.
In short
Google Analytics classifies based on internal rules. Direct, referrers, sessions, and spam create discrepancies. Clean up, then decide.
Frequently Asked Questions
Why don't my numbers match my sales? Due to channel attribution, third-party domain passes, and missing filters.
Should I remove the direct channel? No. Analyze it as a brand indicator without manually correcting it.
Do these pitfalls affect small stores? Yes. An external payment or a blog on a subdomain is enough to create noise.

Enzo
October 28, 2026


