Cart Abandonment Analyzer

diagnose the likely causes of cart abandonment for your specific store and product mix, then build tailored multi-touch recovery sequences across email, SMS, and push notification channels to win back lost revenue systematically.

intent

this skill maps where shoppers bail during checkout, figures out why they're bailing, and builds a sequence of coordinated messages (email, SMS, push) timed to pull them back. use it when your abandonment rate is above 60% and you want to recover revenue without blowing margins with blanket discounts. works for any ecommerce platform (Shopify, WooCommerce, custom). you'll get a ranked diagnosis of friction points plus a calendar of recovery touches with copy, incentives, and send times.

inputs

platform & analytics:

• checkout funnel data (stage-by-stage drop-off %, by device and traffic source)
• customer segmentation data (new vs. returning, LTV, purchase frequency)
• historical abandonment rates by product category, cart value, traffic source
• historical recovery campaign data if you've run prior sequences (open rates, click rates, conversion rates by touch and channel)

external connections:

• ecommerce platform API (Shopify, WooCommerce, custom): needs read access to cart events, checkout funnel, abandoned cart timestamps, customer identity, and order history. set up an API token or OAuth app with scopes: read_orders, read_customers, read_products. store token in ECOMMERCE_API_TOKEN env var.
• email service (Klaviyo, Klaviyo, Braze, etc.): needs ability to create segments, send campaigns, track opens/clicks/conversions, and suppress addresses. OAuth scope: campaigns.write, segments.write, tracking.read. store API key in EMAIL_SERVICE_API_KEY.
• SMS service (Twilio, MessageBird, Klaviyo SMS): needs opt-in list, delivery tracking, and link tracking. requires compliance check (TCPA for US; GDPR for EU). store credentials in SMS_SERVICE_AUTH and SMS_SENDER_ID.
• push notification service (OneSignal, Firebase Cloud Messaging, or native app): optional but recommended for high-value segments. needs device token access and delivery/engagement tracking. store in PUSH_SERVICE_API_KEY.
• analytics or data warehouse (Segment, Mixpanel, custom): optional but strongly recommended for incrementality testing. needs event-level conversion data, cohort isolation, and holdout group analysis.

reference data:

• product catalog with category, price, margin, AOV contribution
• customer acquisition cost by traffic source
• standard shipping thresholds and costs
• payment method availability (credit card, PayPal, Apple Pay, Google Pay, BNPL options available)

procedure

step 1: analyze cart abandonment data

review your checkout funnel to identify exactly where shoppers drop off. pull data for the last 30-90 days. segment by device (mobile, desktop, tablet), traffic source (organic, paid, email, direct), and customer type (new, returning, high-value).

inputs:

• checkout funnel log (timestamps, stage, user ID, session ID)
• cart value at abandonment
• customer segment labels

process:

• calculate abandonment rate at each stage (% who start cart but don't reach next stage)
• flag stages with >30% drop-off as friction points
• cross-tabulate abandonment % by device, source, and customer type
• note any seasonal patterns or product-specific trends

outputs:

• abandonment funnel visualization showing stage, drop-off %, and sample size
• segmented abandonment rates (by device, source, customer type)
• ranked list of top 3-5 friction stages with impact (total $ lost per month)

step 2: diagnose root causes

map the funnel data to likely abandonment reasons. don't guess. look at the evidence.

inputs:

• abandonment funnel from step 1
• checkout page screenshots or heatmaps (if available, shows where users click/scroll)
• customer feedback (support tickets, survey responses mentioning checkout friction)
• competitor checkout analysis (payment options, shipping transparency, trust signals)

process:

• for each friction stage, list 2-3 probable causes based on typical ecommerce drop-off patterns
• cross-reference each cause with supporting evidence (e.g., "shipping page drop-off 25%" + "shipping cost is $5.99" + "free shipping threshold is $75" = likely cause is shipping cost surprise)
• estimate impact of each cause as % of total abandonment
• prioritize causes by combination of likelihood and recovery potential

outputs:

• root cause diagnosis table with stage, probable cause, evidence, and impact %
• for each cause, note which customer segments are most affected (e.g., new customers more price-sensitive; high-cart-value customers hit shipping cost surprise harder)

step 3: design recovery sequence architecture

build the skeleton of your multi-touch sequence: how many touches, which channels, timing between touches, and segmentation rules.

inputs:

• root cause diagnosis from step 2
• product AOV and margin data
• customer segment definitions
• historical recovery data (if available)

process:

• decide on sequence length. for low-AOV impulse products (<$50), run 3-4 touches over 3-5 days. for high-AOV or considered purchases (>$100), run 4-6 touches over 5-7 days. longer sequences increase unsubscribes; too short misses intent windows.
• allocate channels to each touch based on open rates and abandonment stage. email first touch is standard. SMS in middle of sequence (complements email). push notification for returning app users only.
• set timing for each touch. first touch should be 30-60 min for impulse products, 2-4 hours for considered purchases. second touch 6-24 hours. third touch 48 hours. space remaining touches 24+ hours apart.
• define segmentation rules. at minimum: by cart value (under/over $50), by customer type (new/returning), and by abandonment stage. optional: by product category, traffic source, repeat abandoner flag.
• plan content strategy for each touch: reminder (touch 1), social proof (touch 2), objection handling (touch 3), incentive/urgency (touches 4-5).

outputs:

• sequence timeline table with touch number, channel, timing, purpose, and content type
• segmentation decision tree (if cart > $X and customer = new, send full 5-touch sequence; if returning, send 3-touch sequence; etc.)
• channel selection rationale (e.g., "SMS touch 3 because email open rate drops after touch 2")

step 4: write recovery messages

craft copy for each touch in each segment. include subject lines with A/B variants, preview text, body copy, CTAs, and SMS text. every message needs a clear reason to exist (not just "we miss you").

inputs:

• sequence architecture from step 3
• root cause diagnosis from step 2
• product data (images, descriptions, reviews)
• past email performance benchmarks (open rates, click rates)

process:

• for each touch, write a primary subject line and an A/B variant. touch 1 (reminder) might test "Your [Product] is waiting" vs. "[Product] + reviews from real customers". touch 3 (objection handling) might test "We cover shipping" vs. "Free shipping when you complete your order".
• write body copy that addresses the suspected abandonment reason. if root cause is shipping cost surprise, show shipping cost upfront in touch 3. if trust is the issue, add customer testimonials or reviews in touch 2. if comparison shopping, add a unique angle (e.g., "why we're better than [competitor]").
• include product image, price, and cart total so the reminder is specific (not generic).
• craft SMS copy to be under 160 characters with a deep link back to the abandoned cart. test for compliance (TCPA: include clear opt-out language; GDPR: confirm prior consent).
• flag any copy that needs A/B testing due to uncertainty (e.g., social proof vs. discount).

outputs:

• complete message copy for all touches and all segments, including A/B variants
• SMS copy bank (character count verified, TCPA/GDPR compliant)
• list of A/B tests to prioritize (ranked by expected impact)

step 5: define incentive strategy

design an incentive escalation ladder that grows pressure over the sequence without eroding margins. use cart value, customer lifetime value (LTV), and product margin to decide when and how much to offer.

inputs:

• cart value distribution (what % of abandoned carts are <$25, $25-50, $50-100, $100+?)
• product margins by category (COGS % so you know how much discount you can absorb)
• customer LTV data (is this a first-time buyer or a repeat customer?)
• historical discount usage (do customers expect discounts at checkout?)

process:

• set a discount escalation ladder tied to touch number and cart value. example: touch 1-2 = no discount (reminder + social proof). touch 3 = free shipping if applicable, or free shipping offer. touch 4 = percentage off (e.g., 10-15% for carts $50+). touch 5 = dollar amount off or free gift with purchase.
• define exclusion rules to protect margins. never discount customers who already completed a purchase (suppress them from sequence immediately). never discount high-margin, full-price buyers in the last 7 days (train bad behavior). never offer percentage discounts on already-discounted or clearance items.
• set cart value thresholds for incentive eligibility. carts under $20 might get free shipping only (percentage discount too steep on margin). carts $20-50 get 10% off. carts $50+ get 15% off or dollar amount off.
• segment incentive aggressively. returning customers with high LTV don't need discounts; reminder alone may convert. first-time buyers with high cart value should get a strong incentive (they're deciding between you and competitors). repeat abandoners (3+ cart abandons) should be excluded from sequence entirely (deal hunters).
• use unique, single-use discount codes with expiration dates matching sequence timing. expire the code 24 hours after the final touch to create urgency and prevent code sharing on coupon sites.

outputs:

• incentive decision matrix (by cart value, customer type, product category)
• exclusion rules list
• discount code naming and expiration strategy
• projected impact on recovery rate and margin per segment

step 6: set up measurement framework

define KPIs for each touch and the full sequence. plan holdout testing to measure true incrementality (not just attribution). set benchmarks.

inputs:

• historical email/SMS/push performance data
• industry benchmarks (Klaviyo, Littledata, Littledata publish ecommerce benchmarks)
• revenue targets and CAC budget

process:

• define KPIs per touch: delivery rate (should be >95% for email, >95% for SMS to opted-in list), open rate (benchmark 20-35% for cart recovery email), click rate (5-15%), and conversion rate (2-8% for single touch).
• define KPIs for full sequence: recovery rate (% of abandoned carts that convert), revenue recovered ($ value of recovered carts), average order value of recovered carts, repeat purchase rate of recovered customers.
• set incremental KPI: run a holdout test where 10-20% of abandoners get no recovery sequence. measure conversion rate of holdout group and compare to treated group. true incrementality is (treated conversion rate - holdout conversion rate) / treated conversion rate. typical incremental lift is 40-60% (i.e., 60% of attributed recoveries would have happened anyway).
• define measurement dashboard: daily tracking of delivery, opens, clicks, conversions, revenue by touch, segment, and channel. weekly tracking of incrementality test status.
• set a minimum sample size for test confidence: for a sequence with expected 5% recovery rate, you need ~2,000 abandoners per segment to detect a 1% lift with 95% confidence.

outputs:

• KPI definition and benchmark table
• measurement dashboard specification (tools, data sources, refresh cadence)
• holdout test design (sample size, duration, holdout %)
• weekly reporting template

step 7: create optimization roadmap

plan tests for the first 90 days. prioritize based on expected impact and ease of testing.

inputs:

• initial sequence performance data (from step 6 holdout test and week 1-2 live data)
• root cause diagnosis from step 2 (which friction points are most fixable via messaging?)
• copy and incentive variants from steps 4-5

process:

• rank tests by expected impact x ease. high-impact tests: send time optimization (2-3% lift typical), subject line A/B testing (5-10% lift typical), first-touch timing optimization (impulse products show 8-12% lift by moving from 24hr to 1hr). medium-impact: incentive level testing (does 10% beat 15%?), sequence length (4 touches vs. 5?). low-impact: minor copy tweaks.
• plan a testing calendar. week 1-2: measure baseline. week 3-4: test subject lines on touch 1 (quick feedback). week 5-8: test send time optimization (need full sequence data). week 9-12: test incentive levels (high-value segments only to protect margins). week 13+: move to winners, test next hypothesis.
• define win criteria for each test. for subject line test, winner needs 5%+ lift with p < 0.05. for send time, winner needs 2%+ recovery rate lift. for incentive, winner needs to maintain recovery rate while reducing discount depth (margin protection).
• set minimum sample sizes per test. A/B test on subject lines needs ~500 sends per variant. send time test needs 7 days of data (full weekly cycle). incrementality tests need 2-4 week holdout periods.
• document assumptions being tested (e.g., "customers abandon due to shipping cost shock, so showing free shipping in touch 2 will lift recovery by 3%"). after each test, note whether assumption held.

outputs:

• 90-day testing roadmap with test sequence, hypothesis, sample size, and go-live date
• test priority ranking (which tests move the needle most?)
• winning variant documentation process (how do winners get rolled into production?)

decision points

if you have strong analytics data (stage-by-stage funnel, cohort-level conversion data): do the full root cause analysis (step 2). design segmented sequences with stage-specific messaging (e.g., different copy for "abandoned at payment page" vs. "abandoned at cart page").

else if you have limited funnel visibility (only "abandoned" / "not abandoned"): skip step 2 detailed analysis. assume the most common abandonment causes (shipping cost surprise, payment friction, comparison shopping). run a generic 3-4 touch sequence with two-three distinct offers (free shipping, then %, then urgency). measure what actually works, then segment by performance after 2-3 weeks.

if the root cause is checkout friction (missing payment methods, slow checkout, unclear shipping): before building recovery sequences, fix the checkout. adding Apple Pay or Google Pay to payment page may recover more than any email sequence. coordinate with product/engineering. run recovery sequence only after checkout improvements are live.

if you have high new-customer abandonment (>70% of abandoners are first-time buyers): focus recovery sequence on trust-building and education, not discounts. leads: customer testimonials, reviews, guarantee/return policy, product education. reserve discounts for late in sequence (touch 4-5). test "free shipping" vs. "percentage off" because first-time buyers are less brand-loyal.

if you have high returning-customer abandonment: these customers know your brand. they're likely comparing prices or have low intent. try minimal incentive (reminder only, or free shipping for high-cart-value orders). aggressive discounting trains these customers to abandon carts for coupons. measure incrementality closely (holdout test) to avoid wasting email volume.

if you have SMS opt-in rate <20% of your customer base: don't over-rely on SMS. run email-only sequence for non-opted customers. segment your SMS sequence to opted-in users only, and focus SMS touches on high-value carts ($75+) where the ROI justifies the effort of driving opt-ins.

if your product margins are thin (<25% contribution margin): avoid percentage discounts. use free shipping, free gift with purchase, or urgency tactics instead. heavy discounting erodes profitability faster than lost recovery rate.

if abandonment is driven by repeat visitors (customers who abandon multiple times): flag repeat abandoners (3+ cart abandons in 90 days) and exclude them from sequence. they are likely deal-hunters or comparison shoppers, not addressable via messaging. focus sequence budget on first-time and occasional abandoners (higher conversion probability).

if you have no historical recovery sequence data: start conservatively. run 3-touch email sequence (30 min, 24 hr, 72 hr) with minimal discount (free shipping at touch 2, 10% at touch 3). measure performance for 2-3 weeks. once you have baseline, test incrementality with holdout group.

if you have high SMS unsubscribe rates (>2% per campaign): your SMS timing or frequency is too aggressive. space SMS touches further apart (min 24 hrs between touches). reduce total SMS touches in sequence (max 1-2 SMS per sequence). validate opt-in quality (are you SMS-spamming after ambiguous consent?).

if incrementality testing shows <30% true incremental lift (i.e., 70% of recoveries would happen anyway): your sequence is under-optimized or your abandonment recovery naturally high. focus optimization effort on early-touch timing (where intent window is sharpest). test removing low-value touches (if touch 4-5 add little conversion, cut them). consider whether root cause is really addressable via messaging (maybe the issue is checkout friction, not messaging).

output contract

root cause diagnosis table (step 2 output):

• columns: abandonment stage, % of total abandonment, probable cause, supporting evidence, customer segments most affected
• format: CSV or markdown table
• file: [project-name]-root-cause-diagnosis.csv

sequence architecture specification (step 3 output):

• columns: touch number, channel (email/SMS/push), timing (e.g., "30 min after abandonment"), purpose/message type, content description, incentive offered
• format: markdown table or JSON
• file: [project-name]-sequence-architecture.md or .json

message copy bank (step 4 output):

• for each