by Courtney Eckerle
In February, MarketingSherpa covered Eventful's personalization of its weekly newsletter and the launch a new email alert system.
Eventful took home the Best in Show trophy with this campaign by creatively and diligently transforming its email program to highlight relevance in subscriber communications. However, since February, the Eventful team has continued their personalization efforts and added significant dimensions to their email program.
Eventful not only produced an email engagement campaign around personalization, but transformed its entire program by providing relevance when communicating with 21 million subscribers.
In the beginning of this effort, Eventful marketers revolutionized their email product under the pressure of a Google algorithm change that was causing a significant and negative impact to online and mobile traffic.
Throughout the life of this campaign, Eventful marketers and technology teams were "focused on refining the technology and seeing wider distribution of the technology as it gets integrated into more and more product," according to Paul Ramirez, Vice President of Operations, Eventful.
Because so much of Eventful's equity is oriented around live entertainment and leisure activities, Ramirez said it became a high priority of how to recommend more events, while also personalizing further.
"We started thinking about how could we drive greater engagement through richer personalization," Ramirez said. "We looked around and we said well, what are the things that most of our users are doing?"
In answering the question of what Eventful users were doing, music, live entertainment, sports and leisure activities rose to the top, according to Ramirez.
This led Eventful marketers to create an algorithm built from seven years of data on the 21 million subscribers that would power personalized event selection and artist recommendations.
Already launched was a Recommended Performer Alerts product, alerting Eventful subscribers to concerts and other performer-headlined events in their area. That is described in this case study, but focused on is Eventful’s latest efforts for relevancy.
Realizing popular non-performer events were being missed, Eventful began considering other user and subscriber actions that would allow them to recommend local festivals or events that might not have a big name performer.
Data mining from subscribers as well as anonymous site visits, Eventful is now able to send more emails with higher relevancy by recommending non-performer events to subscribers.
Step #1. Build algorithm for increased personalization
This step highlights Eventful’s efforts covered in more detail in the previous case study
, where Eventful launched a new product, its Recommended Performer Alerts, combined with integrating this data into its newsletter.
Ramirez said Eventful has built this algorithm referencing the data store that it has been growing over the past seven years.
"That expanded, essentially, the target that we could aim at to better serve our subscribers in making very smart recommendations," he said.
"So in a nutshell, we have three databases," he added, detailing the sources of data that Eventful drew from for this campaign:
- A consumer database capturing PII (Personally Identifiable Information) and information subscribers tell Eventful about a live event they are interested in
- A database of events featuring their times, locations and performers
- A database housing information about performers
In launching the Recommended Performer Alerts, Ramirez said Eventful set out to merge all of this information for the purpose of personalization, and in the constant effort to make Eventful a better entertainment news source for its community of registered subscribers.
Send out personalized newsletter and alert email
Eventful's Weekly Events Guide showcased a main upcoming event, a sponsored event and recommendation.
Among sections for top venues, trending events and local events, the newsletter encourages the subscriber to "tell us who you like" so they can be alerted when they come to town, as well as adding to the subscriber's recommendations.
Subscribers have the chance to further specify their preferences by clicking an "I'm not interested in [artist's name]" button in each recommendation.
Ramirez said the goal for Eventful is not in the volume of emails sent out, but "matching the preferences of the subscriber for how they want to receive notifications from Eventful."
Step #2. Collect data for non-performer events
After launching the Recommended Performer Alerts, Eventful's team began refining and adding additional data layers to their algorithm. The recommendation engine covered in the previous step was really focused on performer-focused events, such as a Sting concert, for example.
"But, there's a whole slew of events where there is not a known performer," Ramirez said, giving the example of San Diego's Italian festival, Festa.
"Hundreds of thousands of people go because there's a ton of food, they're singing opera in the streets, they reproduce the Mona Lisa on the sidewalk using chalk — it’s a big local event," he said.
However, since there is no specific or well-known performer at Festa, as a recommended event, "[Festa] sort of got missed in the performer recommendation engine that we built, which was around performers and concerts," he said.
According to Ramirez, they had to answer the question, "How do we collect the data about what sort of events [a subscriber] and his family are interested in, when we're really only interested in what kind of concerts he goes to? So, that opened up a whole new world of data possibilities."
When a subscriber is interacting with Eventful.com in some way, either through its webpage or its weekly events guide, they are looking at events other than just concerts or performer events. Building non-performer event recommendation email sends became a matter of harnessing that data.
Also, because so much of Eventful's engagement is done through email — the company sends around a billion and a half pieces of email a year — a lot of the data collected is just waiting to be used.
When a subscriber is on Eventful's page, they can view non-performer events. Ramirez gave the examples of Festa and opening day at the Del Mar Ray racetrack. Data mining those actions became pivotal to this effort.
"We started looking at all of the active and passive activities that a user does in the normal course of either reading an Eventful email, visiting our Web or our mobile site," Ramirez said.
Eventful started looking at more data than the favorites, the demands or the ticket clicks used for the Performer Alerts. In order to drive the intelligence that determines non-performer events to recommend, Eventful began looking at all the commercial activity on the website or within emails taken by its users.
"It sounds like a small thing, but we're talking about hundreds of millions of data points that we've collected," he said.
Those data points are across:
- 21 million subscribers
- 6 million monthly unique visitors
- 5 million average daily site visitors
- Subscriber information, such as favorited performers
All of that, he said, is added to all of the data that Eventful has been collecting against performer events and favorite performers across the last couple of years.
Step #3. Track anonymous site visits
Increasing data for the algorithm has involved bridging the gap between knowing a subscriber's actions, and tracking the data of an anonymous browser coming to Eventful.com or to its mobile sites.
"It is incredibly challenging driving meaningful communication when you don't know who it is you're talking to. So, we've started to tackle that one," Ramirez said.
For example, when someone searches for "San Diego Pet Events" and is led to Eventful, previously that person would only be shown the page for a San Diego Surf Competition, and all the content that wrapped around that page.
Eventful's website would not comprehend the site visitor was interested in pets, and only the most popular events would be suggested if the subscriber left the pet category.
"What we've done today is we've taken [the visitor's] query, 'pets' and deduced that they have a pet," Ramirez said.
All of the stored data about other Eventful users who have pets is used to inform what influences and is displayed on the next click.
Eventful's algorithm draws from seven years' worth of data, taken with its 21 million subscribers and total site actions, without using cookies.
"It's all happening essentially in real time," Ramirez said.
By introducing the personalization algorithms to anonymous site visits, Eventful site visits have increased by 12%, according to Ramirez, and was "easier said than done."
That increase was achieved by not presenting information based on what the team thought, Ramirez said, but on real people and data.
The achievements from this effort are a direct result of the belief that "we serve our customers better by being keenly sensitive to their actions and preferences and by creating products and experiences that match their preferences and anticipate their actions," Ramirez said.
Step #4. Increase sends with increased subscriber relevance
"As a subscription-based email business, we battle the inertia that all emailers battle," Ramirez said. "People tune you out, there's fatigue, they unsubscribe, you don't get delivered. There are all of those normal, day-to-day issues that you face."
For Eventful, personalization is a long-term commitment, and instead of short-term lifts or one particular success metric, they focused on maintaining certain metrics. For example, maintaining open rates even though they were increasing recommendation sends with new non-performer events.
"A year ago, we maybe had the opportunity to send you one or two alerts a month," Ramirez said.
Since an average subscriber has around 30 saved artists, Eventful could only send highly personalized recommendations when one of them is touring and coming to the subscriber's city. Now, it can send more because the frequency of non-performer events is higher.
Ramirez believes email marketers often get too focused on open rates, and can hit a brick wall.
"I do think there’s a practical maximum on what you can expect of an open rate," Ramirez said.
He and his team have been focused on increasing the amount of opens, even if the actual open rate suffers. For example, he would rather a subscriber open two of five email sends rather than opening one of two sends.
"We're able to reach our subscribers more times, more frequently, with meaningful content," he said.
A totally unexpected outcome of the team’s efforts, Ramirez said, is that Eventful's e-commerce with online ticket purchase has increased 66% since the previous case study was published.
Given that non-performer events, which typically do not have online ticket sales, were given equal billing as performer events.
"We hoped, at best, that as people began to engage more with non-performer events that our e-commerce wouldn't suffer," Ramirez said.
But e-commerce has improved, and Ramirez credits being able to engage with customers more frequently, and customers "depending on us as a more comprehensive answer to the question, 'What do you want to do tonight?'"
Since introducing non-performer events into its personalization algorithm, Eventful has seen email opens increase 44%.
"We're reaching them 44% more because we know more about them and we're able to align their preferences with things going on in a way that engages them," Ramirez said.
The flip side of that is, how do you know you're not reaching them too often? From this effort, Eventful also saw its unsubscribe rates decrease by 25%.
"Unsubscribe and complaint is the great proxy for reaching people not enough times with the stuff they want, or reaching them too many times with the stuff they don’t want," he explained.
Finally, by introducing the personalization algorithms to anonymous site visits, Eventful site visits have increased by 12%.
"Driving 12% without spending a dime on advertising is a really cool outcome," Ramirez said. "When we really got serious about personalization, we started talking about it a few years ago and we really got busy a couple of years ago, and now it’s just … it is what we do," Ramirez concluded. To see Eventful marketers speak on relevance and personalization in email, check out MarketingSherpa Email Summit 2014, held February 19-21 in Las Vegas.
Related ResourcesEmail Marketing: Weekly newsletter sees 400% lift in reactivation of inactive subscribers with personalization Email Marketing: BustedTees' personalized send times increase email revenue 8% Email Marketing Relevance: Personalized important occasions campaign by online gift retailer achieves a 39% open rate Email Marketing Segmentation: Clothing brand uses social behavioral data to drive a 141% increase in revenue Email Marketing: Reactivation campaign for performing arts center sees 738% ROI