November 27, 2012
Case Study

Email Marketing: 900% more revenue-per-email from’s automated strategy

SUMMARY: Batch-and-blast is often criticized, but is also widely used. sent generic emails for years, and saw great revenue, but noticed a drop in customer engagement, leading it to dramatically change its approach.

The company overhauled its email strategy and increased revenue-per-email 900% in 12 months, with conversion rates increasing 150%, as well.
by Adam Sutton, Senior Reporter


Nothing seemed wrong with the email strategy at The program had increased readership and revenue for years. However, by 2011, subscriber engagement was down about 25-35%.

"Revenues were growing, but the per-customer effectiveness of the program was falling," says Chris Krohn, President and CMO, generates 70% of its revenue from "certificates," which are discount deals for restaurants. The certificates never expire and can be exchanged across restaurants in the system.

The company grew its email program with nearly constant promotions. The emails were generic, and their frequency had taken a toll on engagement. By 2011, the team sent 22 promotional emails each month, and sold 97% of its top certificates at a discount.

"The only thing that really varied in those 22 emails, by and large, was the level of discount available," Krohn says. "As you jam customers’ inboxes with emails that have no content or particular relevance, open rates, clickthrough rates and conversions start to fall. … It was clear something had to be done."

CAMPAIGN overhauled its email program in 12 short months. The strategy went from "spray and pray," Krohn says, to a sophisticated system of targeted and automated emails. Generic emails were replaced, and results increased across the board. changed its email strategy in six steps:

Step #1. Build internal support

Email is the primary channel for to manage customer relationships and generate revenue. The marketing team planned to change every aspect of its email strategy, and the change would affect nearly every department in the company.

To be successful, the team needed support from the board of directors, as well as these groups in the company:
  • Investors – Significant resources were needed to build infrastructure for the company’s new email program. The team also needed resources to hire staff, since the new email strategy would be far more complex. All of this required support from investors.

  • IT – Data was the lifeblood of the new strategy, and the team needed technical expertise to secure its integrity. IT support was also needed to manage much of the program’s transition, as well as testing of interactive components such as new landing pages.

  • Creative and development – Team members had to design and code new emails and landing pages to match the new strategy and branding standards.

"It isn’t as simple as, ‘give the email team a couple of tools and let them run this,’" Krohn says. "This is something that required top-to-bottom coordination and support."

Step #2. Set a strong foundation’s database was inadequate for the lifecycle strategy it envisioned. The team had to replace several pieces of infrastructure to set a new foundation.

Here are the largest changes:
  • Customer database – Previously, if a customer engaged through a search engine, the company could not tie that information to the customer’s behavior in email or social media. Data for each channel was isolated in siloes.

    In 2011, partnered with a vendor to launch a customer database capable of supporting what Krohn calls a "360-degree view of the customer relationship." The database centralized information on customers’ preferences, demographics and behavior across channels.

  • Cross-channel CRM – On top of the database, the team built a system capable of managing campaigns across channels. This included integration with’s email service provider, as well as a sophisticated analytics system to track results and behavior.

  • Analytical expertise – The team needed skilled hands for the new tools. It built a team of analysts to manage the database and mine it for marketing opportunities. The company grew "from the coordinator level to the vice-president level," says Krohn.

"Now we could really start to reinvent the programs through a set of test-and-control experiments to really start driving improvements."

Step #3. Shift marketing strategy

With new infrastructure in place, changed its marketing strategy in four major ways:

Cut discounts

One of’s best-selling products was a $25 restaurant certificate that cost customers $10. Before the shift, only 3% of the certificates were sold at full price. The team planned to sell more at full price by cutting the number of promotions in half.

Cut promotional email frequency

A direct result of the reduction in discounts was a reduction in the frequency of promotional emails sent to subscribers. The team cut its monthly email frequency from 22 to 11.

Focus on relevance could not cut discounts without offering customers something in return, Krohn says. The team established a variety of triggered and automated emails to reach customers with messages tailored to their behavior. These included:
  • A welcome series – thanked customers, explained the value of, and introduced them to "purchase occasions" for certificates, such as when they traveled or bought gifts

  • Birthday emails – thanked the customer and provided a special offer

  • Certificate use reminder – reached customers who had purchased but not yet used a certificate (more about that campaign in a moment)

Furthermore, emails no longer offered generic, site-wide discounts. Instead, subscribers received offers tied to their behavior and demographics, such as certificates they had purchased and restaurants in their areas.

Set brand standards

The team developed a set of standards to ensure its imagery and messaging were consistent with its strategy. This was the first set of brand standards the company had developed and distributed.

Step #4. Launch automated lifecycle emails now had a wealth of tools and expertise to mine data. One valuable insight identified certificate use as "one of our best predictors of future purchase," Krohn says. In other words, customers who redeemed a certificate were much more likely to buy another.

Further analysis revealed two more insights:
  • Most certificates were redeemed within two weeks after purchase. After that, the team saw a significant drop in redemption rates.

  • Through qualitative research, the team learned many customers who held a certificate for more than two weeks began to question its value. They saw the certificate as "sitting around" and were less likely to buy another in the future.

Certificate reminder email collects revenue when certificates are sold, not when they are redeemed at restaurants. The marketing team hypothesized it could increase the number of repeat purchases by encouraging customers to use the certificates they had purchased but not yet redeemed.

The team created a triggered email to test this hypothesis. Here are its features:
  • Trigger – This email reaches customers who have not redeemed a certificate within 14 days after purchase.

  • Message – The email reminds customers of the "delicious savings" waiting in their accounts. It lists the customer’s recently purchased certificates, including the restaurant name and location. The copy lists three steps to use a certificate, and links to the customer’s account.

  • Frequency – To avoid sending a high number of these emails to customers who purchase several certificates, the team put a cap on the campaign. After the first email, it sends only one email to each qualified customer per month.

"It’s a way to be friendly and remind them, and to help increase their lifetime value. ... We also get revenue out of that, which is not at all the intent of the email, so it’s an added benefit," says Krohn (results metrics are listed below).

Step #5. Predict behavior and trigger emails also automated campaigns to reach customers who perform certain actions. This includes emails that reach customers who:
  • Abandon an item in the shopping cart

  • Purchase a certificate in a new city

  • Participate in a social media contest

"Rather than lifecycle-oriented, they are specific to very particular behaviors," Krohn says. "That makes them feel relevant and specific to customers because we are acknowledging the context of their interactions with our brand."

Automated "due date" email

In one campaign, the team estimates when a customer is "due" to make another purchase. In a hypothetical example, someone who purchased a second certificate after two months and a third certificate after three months might be expected to purchase a fourth within the next three months.

A "due date" is calculated for each customer based on behavioral, demographic and other data. Here’s how the team uses that information to send an automated email:
  • Trigger – An email reaches customers seven days after they have passed their "due date."

  • Message – The email tells customers "your table is waiting" and encourages them to use as a dining guide for a list of occasions. A green call-to-action button urges them to "search & save now!" The bottom of the email lists the names and photos of new restaurants in the customer’s area, and hyperlinks to their landing pages.

  • Frequency – sends no more than one email to each customer every 30 days.

"We were able to prove this one with a hold-out group," says Krohn. "When we send them a predictive ‘next purchase’ message based on the timing and context of what we think they are most likely to purchase, we have much higher order rates overall than those who we hold out of that campaign."

Step #6. Always test and analyze

The team puts every campaign through a test-and-control process before it is launched to the entire database. Tests are performed on a segment-by-segment and campaign-by-campaign basis. This ensures each improves performance instead of hurting it when launched.

The overall hypothesis and strategy of each campaign is tested, as well as its finer points:
  • Images and layout

  • Message content and length

  • Offer placement

  • Level of discount

  • Subject line

  • Timing

"It is not based on a hunch," says Krohn. "It’s literally, ‘two customers have stated a preference for Italian food. Do they respond better when the creative is Italian food, or does it actually not make a difference?’"

Today, the team monitors response to individual campaigns. In the future, Krohn hopes to measure results beyond the campaign-level and determine whether a message brings a net increase in customer lifetime value.


Email is one of’s most important channels. Overhauling its strategy was part of a shift made across the company. Results have improved dramatically, both overall and in the channel.

"It’s been a huge, huge change, and things are going great," says Krohn. "Because the emails are relevant to either [the customers’] lifecycle stage or something very specific in their behavior, we are getting a huge increase in engagement."

Average results for the team’s automated and triggered campaigns:
  • Open rate: nearly 50%

  • Clickthrough rate: 22%

  • Conversion rate: 150% increase over the team’s generic promotions

  • Revenue-per-1000-emails: $200 (a 900% increase over generic promotions)

The team’s standard promotional emails are manually delivered and achieve the following results on average:
  • Open rate: 15%

  • Clickthrough rate: 16%

  • Revenue-per-1000-emails: $20

Discount cut pays off

In 2011, the company sold only 3% of its most popular product at full price. That number increased more than 12-fold.

"Today, 40% of our sales of that product are at the full $10 price," says Krohn.

Opt-outs stay flat cut its number of generic promotions in half, but it also increased its number of automated emails. The result was a drop from 22 emails per month to about 16 to 18.

"The remarkable thing is our customers want to hear from us when we have something relevant to say," says Krohn. "As we have added these triggered and automated campaigns, even when we add more and more of them, we do not see any impact on unsubscribe rates."

Certificate reminder email

The team sees encouraging results from the email that reminds customers to use their certificates. Average results:
  • Open rate: 40%

  • Clickthrough rate: 37%

  • Revenue per 1,000 emails: $11.60 (even though the campaign is intended to increase certificate redemption, not purchases)

Redemption rates for’s certificates have increased by a wide margin. Krohn’s team is convinced this email is a contributor, but he is careful not to confuse correlation and cause.

"We can’t prove chicken or egg just yet," he says. "Usage rates are going up overall right now, and that’s due to a variety of factors, one of which is pulling back on promotional intensity. … Customers use certificates they have rather than stocking up when there is a big promotion."

Creative Samples

  1. Certificate reminder email

  2. "Due date" email


Related Resources

Automated Email Case Study: 175% more revenue and 83% higher conversion rate

Marketing Automation: 25% more engagement, 0% unsubscribe in 4-email series

Email Marketing Research: What information will help you do your job better?

Infographic: Email open rates by time of day

Email Optimization: 72% of marketers test subject lines

Improve Your Marketing

Join our thousands of weekly case study readers.

Enter your email below to receive MarketingSherpa news, updates, and promotions:

Note: Already a subscriber? Want to add a subscription?
Click Here to Manage Subscriptions