The distinction in one paragraph: Added to Cart fires when someone puts an item in the cart; Checkout Started fires when they begin entering their information. They are different Klaviyo events describing different moments of intent, and Klaviyo's own guidance is to run them as two flows: a gentler cart flow for the browsers who never reached checkout, and a firmer checkout flow for the people who almost bought. Most accounts run a single flow off one event, which either misses the cart-only audience entirely or treats window shoppers like near-buyers.
Why the intent gap matters
The global abandonment average sits around 70% of carts, with mobile above 80%, but that number blends two different people. The checkout abandoner selected products, started forms, and stopped: something specific interrupted or worried them, and the recovery job is answering that specific thing fast. The cart abandoner is often still shopping: comparing, saving for later, using the cart as a wishlist. Treating them like a near-buyer (urgency, discounts at hour two) either wastes margin on people who were coming back anyway or pressures people who were three steps earlier in the decision than the copy assumes.
Flow one: Checkout Started, the recovery flow
- Audience: highest purchase intent in the account.
- Timing: fast. First touch within the hour, done within 48 hours for consumables, stretched to 5 to 7 days only for high-AOV considered purchases (that variant is covered in our considered-purchase guide).
- Content: the cart restored, friction answered (shipping, returns, guarantee), then the category's real objection handled, then a structured offer if any. The supplement version of this build is written in full here.
- Offer posture: hold the discount as long as possible; the interrupted convert at full margin.
Flow two: Added to Cart, the nudge flow
- Audience: interested, not yet committed. Filter out anyone who reaches Checkout Started so they hand off cleanly to flow one.
- Timing: slower. First touch at 4 to 24 hours, one or two emails total.
- Content: lighter: the item remembered, a bit of social proof or the answer to the category's comparison question, easy path back. No urgency theater; they were not at the register.
- Offer posture: none, or soft (free-shipping threshold). Discounting people who put one item in a cart trains cart-as-coupon-machine behavior faster than any other mistake.
The deduplication architecture, where accounts break
The failure mode is a customer qualifying for both flows and getting two competing sequences about the same product. The rules that prevent it:
- Checkout Started wins. Flow two's filters exclude anyone with a Checkout Started event in the window, and Klaviyo's flow filters plus a has-not-started-checkout condition handle the routing.
- One source of truth per intent tier: most Shopify accounts should trigger checkout recovery on the Shopify Checkout Started event, and cart nudges on Klaviyo's Added to Cart, with the onsite tracking snippet installed so the event actually fires.
- Both exit on purchase, instantly.
- Smart Sending off for recovery flows, deliberately: a recovery email suppressed because a campaign went out yesterday is a recovered order lost. Manage overlap with the exclusions, not the throttle.
One more tier below: browse abandonment
Below Added to Cart sits Viewed Product, the browse abandonment audience, largest of the three and gentlest to handle. In considered categories it is where most intent lives, and it gets its own build in our browse abandonment guide. The three tiers together (browse, cart, checkout) form an intent ladder, with message pressure scaled to rung.
What to measure, per flow separately
- Recovery rate and revenue per recipient for each flow on its own: blending them hides whether the cart flow adds anything beyond what checkout recovery already catches
- Discount share of recovered revenue per flow: the cart flow's should be near zero
- Handoff integrity: spot-check that cart-flow members who start checkout actually exit into the checkout flow rather than riding both
- Clicks, not opens, as everywhere (Apple MPP inflates opens 15 to 35%)
Frequently asked questions
Do small brands need both flows?
Start with Checkout Started recovery: highest intent, highest return. Add the cart flow once the first is tuned; add browse abandonment after that. The ladder gets built top down.
Which event should trigger checkout recovery on Shopify?
Shopify's Checkout Started, as the single source of truth, deduplicated per session. Mixing checkout triggers from multiple integrations is how double-sends happen.
Why is my abandoned cart flow not triggering for everyone?
Added to Cart requires Klaviyo's onsite tracking active and the visitor to be cookied (a known profile). Anonymous visitors do not enter; that is normal, and it is why the cart audience is always smaller than analytics suggests.
Should the two flows share creative?
Share the design system, not the copy. The checkout flow answers objections; the cart flow re-warms interest. Same template family, different jobs.
Sources
- Klaviyo Community. Abandoned cart vs Checkout Started guidance.
- Klaviyo Help Center. Creating an abandoned cart flow.
- Mailmend. Cart abandonment statistics, Baymard meta-average.