How Much of Your Cart Abandonment Is Bots

SummarizeChatGPTPerplexityGeminiClaude

If you're reading your abandonment rate off a plugin or your Shopify analytics tab, some share of what you're looking at isn't a shopper at all. It's automated traffic — and depending on the day, it can be more than half your sessions.

The traffic your dashboard can't tell apart

A storefront receives requests from more than shoppers. Search engines crawl it to index pages. Uptime monitors ping it to confirm it's alive. Comparison and price-tracking tools revisit product pages on a schedule. Ad platforms and social apps fetch pages to generate link previews. Scrapers and automated browsing agents increasingly load pages the way a person would — executing JavaScript, even adding items to cart — in ways that look, to a basic analytics tool, indistinguishable from a real visit.

None of that traffic is a customer deciding whether to buy. All of it can register as a session, and depending on what triggered it, as a cart.

Why this specifically wrecks abandonment metrics

Abandonment metrics are a ratio: carts that don't convert, divided by carts created. Automated traffic inflates the denominator without ever being a candidate to convert. That has two compounding effects:

• Your abandonment rate looks worse than it is. More non-human carts push the rate up even though nothing changed about real shopper behavior.
• "Lost revenue" estimates get inflated further. Multiply an inflated abandonment rate by average order value to estimate what abandonment is "costing" you, and the bot share flows straight into that number too, compounding the distortion.

Measured across our fleet, bot sessions have run 27% to 60% of daily sessions on some days (measured 2026-08-25) — that's not a rounding error. It's the difference between a healthy funnel and an alarming one, read off the exact same raw data.

What to do about it

The fix isn't to try to eyeball which sessions look automated — most tooling that reports session and cart counts has no way to separate the two, and a merchant staring at a spreadsheet can't either. The fix is to make sure whatever measures abandonment value for you has already done that separation before the number reaches you.

Navona excludes anomalous carts from Abandoned Cart Value — the dollar figure that represents what's actually at stake — while still counting them in raw cart counts, so the volume hitting your store stays visible and the revenue figure isn't inflated by traffic that was never going to buy. That split matters: hiding the bot traffic entirely would make it look like it isn't happening, and letting it into the revenue number would overstate what's really on the table.

If a tool you use reports a single "abandoned cart value" without saying anything about how it handles non-human traffic, ask. The gap between a properly filtered number and a raw one widens as bot traffic grows — and it's growing across the whole web, not just on Shopify.

The honest takeaway

A raw session count, a raw cart count, and a raw abandonment rate all carry an unknown, often large, share of non-human traffic. Treat any of those numbers — from any tool, including ours — as directional rather than exact unless you know it's been adjusted. We wrote separately about the same discipline applied to attribution numbers: a correctly-computed figure can still mislead if you don't know what it's actually counting.