It’s common for businesses to notice differences between the number of purchases or transactions reported in GA4 and the numbers recorded in back-end systems. While it might seem alarming at first, these discrepancies are typically the result of how data is collected, processed, and attributed across platforms. Here are some of the main reasons why the numbers might not match exactly:

1. Tracking Issues and Data Loss 
   GA4 uses client-side tracking, capturing data via browser events or network requests, even if using a server-side GTM collection method. As a result, issues like browser crashes, JavaScript being disabled, ad blockers, lost connections, or rejected cookie consent can prevent purchases from being recorded in GA4, even if the back end logs them.
2. Differences in Validation Logic
   The back-end system typically records only "valid" or fully completed transactions, for example, after payment has been authorised. In contrast, GA4 records whatever purchase event is fired on the site, regardless of whether the payment was ultimately successful, refunded, or cancelled. If GA4 events are triggered too early in the purchase process (i.e. before full payment confirmation), this can cause inflated numbers compared to the back end. This can typically be remedied by validating the implementation of theGA4 purchase event, ensuring that it only fires after the payment has been authorised (i.e. on a payment confirmation page).
3. Timing and Data Processing Lags
   GA4 data is often processed with some delay, especially for high-traffic websites. Additionally, users might complete a transaction but take longer to trigger events that GA4 picks up (for example, lingering on a confirmation page). Back-end data, on the other hand, typically reflects real-time, finalised transactions.
4. Duplicated or Missed Events
   If event tagging isn’t properly configured, GA4 might record duplicate purchases (for example, if users refresh the confirmation page) or miss purchases entirely if the tracking code fails to fire. Proper event de-duplication and robust tagging strategies are essential to minimise these discrepancies.

When businesses start using BigQuery exports from GA4 for deeper analysis, they often notice that the numbers don’t exactly match what’s shown in the GA4 reporting interface. Although the data is connected, there are several reasons why discrepancies can occur between GA4 reports and BigQuery queries:

1. Processed vs. Raw Data 
   The GA4 interface presents processed data. It applies filters, event-counting logic, attribution settings, and sometimes even modelling (such as for consent mode or data thresholds). BigQuery exports, on the other hand, provide raw, unprocessed event-level data. This means that without additional transformations, calculations, or filtering, the two sets of data may tell different stories.
2. Differences in Sampling and Thresholding
   In the GA4 interface, Google may apply sampling to large datasets or thresholding to protect user privacy, especially for reports with small user counts or sensitive dimensions. These measures can suppress or modify the data shown. In BigQuery, it's possible to access to the underlying events as they were logged, without these adjustments (unless manually applied).
3. Timing Differences and Data Freshness
   GA4 reporting and BigQuery exports may not update at the same time. While GA4 reports might refresh more frequently for high-priority dashboards, BigQuery exports are typically processed in batches, often once a day (depending on the type of export configured). As a result, when comparing metrics in close-to-real-time, discrepancies can appear simply due to timing.
4. Query Logic Differences
   The way queries are written in BigQuery can greatly impact results. For example, if the query doesn't correctly account for session starts, user deduplication, or event filters (such as ignoring non-conversion events), the BigQuery numbers may differ from what GA4 is showing natively in-platform, where all that logic is already built-in and accounted for.
5. Attribution Settings
   In GA4 reports, attribution models (such as data-driven attribution or last click) are automatically applied when calculating conversion metrics. In BigQuery, attribution logic needs to be manually rebuilt for the analysis to mirror GA4’s settings. Otherwise, the way conversions are associated to channels, campaigns, or users may differ. 
6. HyperLogLog++ Algorithm
   Measuring exact distinct counts (i.e. cardinality) for large datasets requires significant memory and affects performance. GA4 properties use the HyperLogLog++ (HLL++) algorithm to estimate cardinality for most used metrics including “Active Users” and “Sessions”. Discrepancies can be caused when comparing GA4 and BigQuery data due to the use of HLL++ in GA4, while not replicating the same counting behaviour in BigQuery. These discrepancies can be avoided, however, by leveraging the same HLL++ algorithm for such common metrics on BigQuery data.

When it comes to analysing web and app activity in Google Analytics 4 (GA4), one of the most common pitfalls is trying to report on “today's” data. While GA4 provides real-time reporting and updates data throughout the day, relying on in-progress data for decision-making or performance reporting can lead to serious misinterpretations. Here’s why:

1. Data Processing Delays 
   Unlike real-time reports (which show a basic snapshot of active users and events), standard GA4 reports, and custom Explorations depend on processed data, and that processing isn't instantaneous. It can take hours for GA4 to fully ingest, model, and attribute event data, meaning that numbers for the current day are often incomplete or under-reported.
2. Event Deduplication and Attribution Happen Later
   GA4 applies deduplication rules, attribution modelling, and session stitching as part of its processes. These steps ensure that sessions and conversions are counted correctly, but they aren’t immediate. When viewing today's data, it may not reflect final attribution paths or properly filtered event counts, leading to inflated or missing data.
3. Real-Time vs. Standard Reports Confusion
   Real-time data and standard reporting in GA4 use different pipelines. Real-time data focuses on events occurring within the last 30 minutes and is designed for monitoring activity, not for accurate performance reporting. When pulling numbers from real-time views and comparing them to the standard reports, there may be inconsistencies, especially for metrics like conversions, revenue, and engagement time.
4. Sampling and Thresholds May Apply More Aggressively
   When datasets are small or incomplete (as they often are during the day), GA4 may apply sampling or privacy thresholds even more conservatively. This means that some dimensions or metrics might be withheld or generalised in reporting until more data accumulates.
5. Back-End Systems Might Not Sync Yet
   When importing offline conversions or using back-end systems that sync data into GA4, those imports often occur on a scheduled basis, sometimes overnight. As a result, important events like purchases, leads, or user property updates may not show up in today's GA4 reports until much later.

One of GA4’s more powerful—but frequently misunderstood—features is its ability to create sequential user segments. These segments define conditions based on the order in which users take specific actions (for example: User viewed a product page, then added the product to cart, then made a purchase). This is incredibly useful for understanding user journeys—but it’s critical to understand how GA4 applies these segments when reports are viewed.

A common pitfall is assuming that applying a sequential user segment, GA4 will only show sessions or events that match the exact sequence. In practice, GA4 will in-fact show all data for users who ever matched that sequence within the selected date range, not just the session in which the sequence occurred.

For example, consider a segment defined for users who performed the following sequence:

1. Viewed a product

2. Added product to cart

3. Made a purchase

    

If a user completed that sequence on Tuesday, and then returned on Friday to browse other unrelated products without purchasing, the segment will include data from both sessions—because that user matched the defined behaviour at least once during the selected report date range (assuming the report date range is, for instance, Monday to Sunday).

This behaviour can lead to confusing or inflated results when the expectation is that the report will reflect only the sessions or events that match the sequence. Instead, the report includes all user activity, regardless of whether subsequent sessions continued the same pattern.

Here are some tips for using sequential segments effectively:

1. Use Filters in Combination with Segments 
   To focus only on the session where the sequence occurred, filter by session-level conditions or use additional logic in Explorations.
2. Be mindful of data scope
   Remember that user segments operate at the user level, not the session or event level. For more precise control, consider building session-based segments instead.
3. Document segments clearly
   When sharing reports with stakeholders, clarify what a segment includes and how GA4 is applying it—this helps avoid misinterpretation of the data.

GA4 offers a flexible, future-focused approach to digital analytics—but that flexibility comes with a learning curve. Many of the most common pitfalls don’t stem from bugs or broken tracking, but from key differences in how GA4 processes, attributes, and presents data compared to other tools in the market. Whether comparing "Key Event" counts across platforms, mixing dimensions/metric scopes in reports, or finding discrepancies between GA4 and BigQuery, it's essential to approach analysis with a clear understanding of how the data is structured at a foundational level and what each number really represents. 

By being aware of these common traps and proactively aligning your implementation and reporting strategies, you can avoid misinterpretations and make smarter, data-driven decisions. As GA4 continues to evolve, staying informed and intentional in your analytics setup will be key to making the most of what it has to offer. 

If you're unsure about your setup or are seeing inconsistencies in your reports, our team is here to help. We offer expert audits, implementation support, and custom training to ensure you're getting clean, actionable insights from your data. Reach out today to start making GA4 work for you.