Raiders of the Lost Cart How APMEX mined a hidden treasure from abandoned orders and captured 13% of lost revenue David McCarty Vice President of Internet Marketing The American Precious Metals Exchange
Case Study Background
Case Study Background • The Company : The American Precious Metals Exchange • The Marketplace: Online precious metals investments • The Product: Investment grade gold, silver, platinum and palladium coins, rounds and bars • The Channel : APMEX.com online shopping cart system • The Campaign Objectives : – To bring shoppers back to website – Revisit abandoned shopping carts – Reiterate and clarify value proposition – Connect troubled shoppers with customer service 3
Case Study Background • Target Audience : – Primary demographic is Male, 40-65, College/Some College, Professional, HHI of $100K, $250K Investments – New and Existing logged in users with open shopping carts – Known animosity to intrusive techniques • Method: Triggered email series • Key Performance Indicators: – Open rate – Click through rate – Conversion rate – Recovery rate – Recovered revenue 4
Case Study Background • Metrics: – Emails received • Received 1st email (# of customers) • Received 2nd email • Received 3rd email – Emails opened • Opened 1st email • Opened 2nd email • Opened 3rd email – Emails clicked (for each 1st, 2nd, and 3rd email) – Purchases (for each 1st, 2nd, and 3rd email) – Revenue (for each 1st, 2nd, and 3rd email) – Funnel data (for each 1st, 2nd, and 3rd email) 5
Case Study Background • Goal: To recover customers that abandon the site within the checkout process • The Approach: – Send all abandoned customers a sequence of three recovery emails – one after 24 hours, a second after 48 hours, and a third after 7 days – Send half of all abandoned customers text emails and the other half HTML emails • Research Questions : – What percentage of customer orders will be recovered in a set time period (one month)? – Which treatment (email type) will produce the highest recovery rate? – Which treatment (email type) will produce the highest revenue per order? 6
Case Study Background Additional Observations : – This test would help us understand how customers respond to a customer service initiated email effort on a regular basis – This test would also help us understand how customers respond to HTML style in email 7
Email Frequency Analysis • Data Analysis Review* Visits to Purchase – Visits to purchase 1 visit (% of All Transactions) • 40% purchase on the first visit 2 visits 2% 1% 3 visits • Multiple visits per day to 4 visits 4% purchase 6% 5 visits 6 visits – Days to purchase 7% 7 visits 40% • 54% of customers buy the 8 visits 7% 9-14 visits same day 2% 15-25 visits • Only 17% of customers buy 2% 26-50 visits 3% 3% between 1 and 7 days, with 51-100 visits 5% 101-200 visits the majority buying same day 7% 11% 201+ visits *Source: MarketingExperiments Data Analysis Document, May 15, 2009 8
Email Frequency Analysis • Key Considerations* Days to Purchase – The objective of basket recovery (% of All Transactions) is to bring customers back that 2% 0% are not already intending to do 0 days 5% so as quickly as possible 1 days 7% 2 days – Basket recovery best practices 3 days 8% 4 days call for an immediate/1-hour 5 days send after cart abandonment 6 days 7% 54% 7 days – Visitors that are already 1% 8-14 days 2% 15-30 days accustomed to returning same- 31-60 days 2% 61-120 days day to make a purchase may 2% 2% 4% 121-365 days 3% feel spammed 366+ days *Source: MarketingExperiments Data Analysis Document, May 15, 2009 9
Email Frequency Design • Recovery Email Scenarios – The customer registers a return visit before the first email • Basket recovery clock will reset – If a customer’s cart changed after receiving a basket recovery email, and abandonment follows • Basket recovery will continue with new cart – The customer receives all three basket recovery emails • Basket recovery stops – The customer replies to an actual customer service representative (by phone or email) • The customer service representative will flag their account immediately to stop basket recovery – One or more items in the customer’s abandoned cart are out of stock • Basket recovery pauses until items are back in stock 10
Test Design Summary • Three Email Sequence – The first email is sent 24 hours after an incomplete order; the second email sent 48 hours after initial cart abandonment, and the third email sent 7 days after initial abandonment Email #1 24 hours Email #2 48 hours Email #3 7-days Cart Cart Contents Contents 11
Strategy & Tactics
Strategy & Tactics • Strategy #1: Plain Text Format – Test against the effects of HTML formatting on customers (size, color, shape) – Mitigate possible constraints created by HTML formatting in email content • Most SPAM is also HTML- formatted • HTML emails are rejected by SPAM filters twice as often as text-based messages • HTML takes more resources and bandwidth than text • HTML doesn’t always render properly • HTML is prone to delivery problems 13
Strategy & Tactics • Strategy #2: HTML Format – Test the effects of HTML formatting on customers (size, color, shape) – Pictures of items from abandoned cart included • Header and footer that match existing email formatting • Separate font style for cart items – Gives the customer more options to return to APMEX.com (footer) 14
Strategy & Tactics • Tactic #1 – Email Cadence – 1st email 24 hours after abandonment • Allows users to come back naturally to check prices and/or competition • Mitigates against customer tension created by interrupting the natural buying cycle – 2 nd email 48 hours after abandonment • Allows APMEX.com to increase sales 17% during days 1 to 7* – Final email 7 days after abandonment • Increases urgency with “Last Chance” messaging *Source: MarketingExperiments Data Analysis Document, May 15, 2009 15
Strategy & Tactics • Tactic #2 - Recovery Email Header – Email comes from a realistic persona • Specific name and title • Specific email address – Subject line is relevant to visitor’s recent activity Jana Hale, Customer Service Manager [jana@apmex.com] Joe Smith [jsmith223@gmail.com] Your Recent Purchase Attempt at APMEX.com 16
Strategy & Tactics • Tactic #3: Content – 24-Hour Email • Customer service tone and personal signature • Listed cart items • Reminder and curiosity establish a need to continue – “View updated prices...” • Descriptive underlined links make it easier to bring customers back where they want to go 17
Strategy & Tactics • Tactic #3: Content – 24-Hour Email • Value of APMEX.com re- emphasized – Quantitative statements – Logic built-in to customize statements by cart contents • Concerns are mitigated with phone number and support page • The P.S. highlights the • 1,245 Bullion Products to choose from strongest point of value • Competitive prices with no hidden fees • Industrial strength secure shopping • Bullion experts to answer all your product questions 18
Strategy & Tactics • Tactic #3: Content – 48-Hour Email • Specific mention of time frame of emails • Adjusted content and presentation for originality • Increase curiosity to see “updated prices” and call to action to “continue” • Consistent tone and contact • Reiteration of strong value statement in the P.S. 19
Strategy & Tactics • Tactic #3: Content – 7- Day Email • Tells the customer this is the last email they will receive for this cart • The body copy focuses only on the strongest point of value • Reiterates urgency in P.S. 20
Strategy & Tactics HTML Email Text Formatted Email 21
Initial Test Results: 8/7/09 -9/7/09
Results Recovery Efforts % Orders % Revenue Rev/Order All Emails 16.83% 8.28% $1,615 TEXT 8.09% 4.47% $1,813 HTML 8.74% 3.81% $1,431 Rel. Difference (text vs. HTML): -7.36% 17.40% 26.73% In Summary: This test recovered nearly 17% of abandoned orders in the checkout process and 27% 8.28% of total lost revenue. Customers receiving MORE REVENUE text emails recovered $382.43 more revenue per FROM TEXT EMAIL abandoned customer than HTML emails. 23
Results Recovery Efforts % Orders % Revenue Rev/Order All Emails 16.83% 8.28% $1,615 TEXT 8.09% 4.47% $1,813 HTML 8.74% 3.81% $1,431 Rel. Difference (text vs. HTML): -7.36% 17.40% 26.73% Key Point: Customers with smaller carts responded more favorably to basket recovery efforts. Customers with slightly larger carts respond more favorably to text emails, while customers receiving HTML emails respond slightly more often (but with smaller carts). 24
Continuation Results: 2/1/11 – 2/28/11
Results Recovery Efforts % Orders % Revenue Rev/Order All Emails 3.60% 12.98% $3,254 TEXT 1.87% 7.68% $3,714 HTML 1.74% 5.31% $2,759 Rel. Difference (text vs. HTML): 7.45% 44.65% 34.63% In Summary: Unaltered continuation of test protocol has continually recovered nearly 4% of 35% all abandoned customers in the checkout process MORE REVENUE recapturing nearly 13% of total abandoned FROM TEXT EMAIL revenue. 26
Recommend
More recommend
