Why some stats look inaccurate

If a number on your dashboard or analytics page looks too high, too low, or just wrong, it is usually one of a few known, explainable reasons rather than a bug. This article walks through the most common ones so you can read your stats with confidence.

Visitor tracking is newer than your orders

The biggest cause of surprising rates is timing. Bundlex has tracked orders since launch, but visitor and impression data only starts on April 12, 2026. That means any shop installed before then has months of order history with no matching visitor data.

Because rates divide one number by another, mixing complete order data with incomplete visitor data produces inflated results. The affected metrics are:

• Cart rate (add-to-cart visitors / visitors)
• Conversion rate (orders / visitors)
• Revenue per visitor (total revenue / visitors)

For a shop that has been live a while, all-time orders can easily exceed all-time visitors, which pushes conversion rate to 100% (it is clamped) and makes revenue per visitor look enormous. The dashboard discounts table always shows all-time figures with no date filter, so these columns are the most likely to look off.

How to see accurate rates

The fix is to look at a window that sits entirely after tracking began.

1. Open the Analytics page, which has a date picker. The dashboard discounts table does not.
2. Pick a date range that starts on or after April 12, 2026. Avoid "All time" for rate questions.
3. Read your cart rate, conversion rate, and revenue per visitor from that range.

With a range fully inside the tracked period, every analytics metric lines up correctly.

What is always correct

Order-based numbers never had a tracking gap, so these stay accurate even on "All time":

• Orders (synced from Shopify, full history)
• Added revenue (the upsell delta, from order line items)
• Total revenue (from order data)
• AOV (total revenue / orders, both complete)

Why cart rate can be lower than conversion rate

It looks backwards. In a normal funnel, more people add to cart than actually buy, so cart rate should be higher than conversion rate. In Bundlex the opposite is common, and it is not a bug.

The reason is that the two numbers count different things:

• Cart rate only counts add-to-cart events that happen through the Bundlex widget form on the product page.
• Conversion rate uses the server-side order count, which captures every purchase path.

Plenty of customers reach checkout without ever touching the widget form, for example through a quick-buy or Buy Now button, a collection page, search results, a re-order from order history, or with an ad blocker that stops the widget's JavaScript. Those orders are still attributed and counted, but the add-to-cart event was never recorded. So conversion rate captures the full set while cart rate captures only a subset, and conversion rate above cart rate is expected.

For the same reason, the conversion funnel on the analytics page is not a strict funnel. The "Added to cart" step can show fewer events than "Orders," so the bars may not always step down. Treat add-to-cart and order counts as complementary signals rather than two stages of one funnel.

What a dash means versus a zero

These two look similar in the discounts table but mean different things:

• A dash (-) means there is no data for that metric yet, for example a brand-new bundle that has not collected any visitors.
• A zero (0) means there is data and the value is genuinely zero.

So a dash in a rate column is not a value of zero. It just means the metric has nothing to calculate from yet.

Total Revenue can look inflated across bundles

If a single order contains items from two different bundles, each of those bundles claims the full order total in its Total Revenue, AOV, and Rev/Visitor columns. The same order is counted in both rows, so adding up the Total Revenue column across bundles can come out higher than your real revenue.

This is expected behavior, not a bug, and in practice it affects only a small share of orders (around 1.3%). Added revenue is not affected, because it is calculated per line item and splits cleanly between the bundles. If you want a clean shop-wide total, rely on the shop-level summary cards, which count each order only once.

New bundles are accurate from day one

Any bundle created after visitor tracking went live collects orders and visitors together from the start, so its rates are correct right away. The pre-April gap only affects bundles that existed before tracking began, and it keeps shrinking over time as the tracked period grows relative to the untracked one.

Quick rule of thumb: if a rate looks wrong, check the bundle's age and the date range. Pre-April 2026 bundles viewed on "All time" are the usual culprit. Pick a range after the tracking start date and the numbers will make sense.