70.2% of online shopping carts are abandoned before checkout finishes, per Baymard Institute's rolling meta-analysis of 50 studies. That number gets repeated everywhere, usually followed by a stat that doesn't hold up — a made-up-sounding recovery rate, a suspiciously precise "send at exactly 1 hour" rule, or a WhatsApp conversion figure nobody can actually source.
This piece sticks to what we could independently verify: Baymard's own reasons data, Klaviyo's published benchmark numbers, and the honest gaps where the popular stat turned out to be unsupported. Where the evidence runs out, we say so instead of inventing a number — and we tell you what to do anyway.
Key takeaways
- 70.2% of carts are abandoned (Baymard, 50-study meta-analysis, updated Sept 2025) — a rolling average, not a fixed fact, and it has crept up as Baymard added studies.
- The single biggest verified reason really is cost surprises — 39% blame extra costs like shipping, tax or fees showing up late. But four other reasons (slow delivery, distrust with card details, forced accounts, a dragging checkout) are about trust and effort, not price, and together they add up to more.
- A cart-recovery email earns real money: ~50.5% open, ~6.25% click, ~3.33% place an order (Klaviyo), and automated flows earn roughly 30x the revenue per recipient of a one-off campaign.
- The famous "1 hour vs 24 hour" timing statistic traces back to an unverifiable 2016 press release — don't build a strategy on it. Send while intent is warm; that's the part that's actually defensible.
- "Texting converts better than email" doesn't hold up in the newest aggregate data either. The real win is sequencing a fast channel with email, not picking a winner.
How many carts are actually abandoned?
The number you'll see cited everywhere is Baymard Institute's — and for good reason: it's a continuously updated meta-analysis of dozens of individual studies, not a one-off survey with an agenda.
- It's a rolling figure, not a fixed constant — Baymard's tally has climbed as more studies were folded in (an earlier version of the same analysis, based on 14 studies, put it at 65.2%).
- That means the honest move is to cite the live Baymard page, not a screenshot of this number from a 2023 blog post.
- Whatever your own number is, 70% is the baseline everyone is compared against — it's not an alarming outlier, it's the default.
70.2%
of online shopping carts are abandoned before checkout completes.
Source: Baymard Institute, meta-analysis of 50 studies, last updated September 2025
Why do shoppers actually abandon their carts?
The reason you'll see quoted everywhere turns out to be real: 39% of shoppers who abandon mid-checkout blame extra costs — shipping, tax or fees that show up too late. That figure is straight off Baymard's own live tracking page, not a recycled blog estimate. But cost isn't the whole story:
Shoppers can cite more than one reason, so these don't add up to 100%. The pattern that matters: four of the five verified reasons have nothing to do with price — they're about trust and effort. Fix transparency and friction, and don't assume a discount is the only lever.
Does a recovery email actually make money?
Yes — and the funnel shape tells you exactly where it earns that money. Klaviyo's benchmark report (based on 2023 send data across 143,000+ flows, published May 2024) is the most complete published dataset we could verify:
30x
the revenue per recipient an automated cart-recovery flow earns compared to a one-off campaign blast ($3.07 vs $0.10 RPR).
Source: Klaviyo 2025 Benchmark Report, aggregated ecommerce industries
This is Klaviyo's own customer data, not a universal law of ecommerce — treat it as a credible reference point to set expectations, not a guarantee for your list.
How fast should the first recovery message go out?
You'll find a very specific, very quotable statistic on this: send within 1 hour for a 20.3% conversion rate, versus 12.2% if you wait a day. We tried to verify it — it traces back to a single 2016 press release, and it didn't hold up in this round of checking. Don't build a timing strategy on a decade-old number nobody can re-source.
What is defensible is the underlying logic, not the decimal points: interest decays. A shopper who was one tap from paying twenty minutes ago is a warmer lead than the same shopper a week later. So the safe, evidence-consistent version of the rule is directional, not precise:
- 1Fire the first nudge the same day, while the cart and the intent are both still fresh.
- 2Follow up the next day if there's no click — this is where you add detail, not just a reminder.
- 3Send a final, slightly more generous message a few days out if margin allows, then stop.
- 4Suppress the whole sequence the instant they buy — nothing kills trust like a "still thinking it over?" email for an order that already shipped.
Does WhatsApp actually beat email for cart recovery?
Honestly: the cart-recovery-specific benchmark data for WhatsApp isn't published anywhere we could verify — it's a newer channel for this use case and the rigorous studies haven't caught up yet. Anyone quoting you an exact WhatsApp abandoned-cart conversion rate is quoting a number nobody can source.
What we could verify is the closest available proxy: aggregated revenue-per-recipient for SMS versus email. And it undercuts the usual "texting just converts better" pitch.
Average performer
Top 10% performer
Source: Klaviyo 2025 Benchmark Report. SMS, not WhatsApp — we're citing it as the closest verifiable proxy, not a stand-in claim. Note the reversal: among top performers, email pulls ahead of texting. Channel alone isn't the lever; relevance and timing are.
What does a combined WhatsApp + email sequence actually look like?
Given the data above, the honest recommendation isn't "WhatsApp converts X% better" — it's that a fast channel and a detailed channel do different jobs, and sequencing them beats picking one.
- Write to the verified objections, not a generic nudge: show the total cost upfront, a real shipping estimate, or a guest-checkout option — that answers three of the five verified reasons people actually leave (cost surprises, delivery expectations, forced signup).
- Don't run both channels in parallel forever — once someone engages on one, drop them from the other so it reads as one conversation, not a chase.
- Kill the sequence the moment the order completes. This should be an automatic trigger, not a manual step someone forgets.
| Step | Timing | Channel | Job |
|---|---|---|---|
| 1. Fast nudge | Same day | Short, low-friction reminder — it's read in minutes, not hours | |
| 2. Detail follow-up | Next day, if unopened | Show the cart, address a real objection (cost, trust, effort) | |
| 3. Last call | A few days out | Whichever channel they've engaged | Slightly better offer, then stop for good |
Frequently asked questions
- What's a realistic abandoned-cart recovery rate?
- Be skeptical of any single "recovery rate" quoted without a source — most of the specific percentages circulating didn't hold up when we checked them. The closest verified reference point is Klaviyo's data: about 3.33% of people who receive a cart-recovery email go on to place an order. Your real number depends heavily on traffic quality and how close to purchase the shopper actually was.
- Do I need to offer a discount to win the sale back?
- Not as a blanket first move. The single biggest verified reason is a cost surprise — 39% blame extra shipping, tax or fees that show up late — so showing the full cost earlier fixes more of that than a coupon does. The next four reasons skew toward trust and effort, not price. Fix transparency and friction before defaulting to a discount; save the incentive for the final message in the sequence, if margin allows.
- Should cart recovery run on WhatsApp, email, or both?
- Both, sequenced rather than duplicated. Use WhatsApp for the fast, low-friction first nudge — it's read in minutes — and email for the detailed follow-up that addresses a specific objection. There's no verified data showing WhatsApp alone converts better for cart recovery specifically; the win comes from the sequence, not a single "best" channel.
