When Victor Cho interviewed for the CEO position at Evite years ago, he was asked whether he thought the company would be another startup flop like MySpace or a business that could be of value. At the time, Evite was struggling from declining user growth after facing competition from Facebook, which had a deeply embedded event function, as well as scrappier startups such as Punchbowl. In a 2010 article, tech blog TechCrunch opined that “Evite sucks,” calling it the “MySpace of online invitation services” with a site design that was circa-1998.
After looking at some company data, Cho quickly realized that “this is actually very reparable,” he said. “The message I gave to them was that probably within two years, I should be able to stabilize the business.” How could he be so confident? “The reason I knew that was because I went through the customer experience a little in-depth and I saw how bad it was,” he said. And, also importantly, “the slope of degradation was very low for a business that had neglected its customer experience.”
In 2014, he took the job. The decision was “based securely on my belief that bad customer experience was the gating factor that was dragging the system down,” said Cho at the recent Wharton Customer Analytics Initiative conference. He calls Evite a “tech, anti-tech business.” While it uses AI and software, its mission is to foster personal contact. “Our purpose as a business is much broader than being the world’s biggest invitation service,” he said. “It’s connecting people together face-to-face.”
Today, Los Angeles-based Evite has reinvented its site and serves more than 100 million annual users. It just reached its two-billionth invitation milestone and the company is celebrating its 20th anniversary this year. Evite also has doubled its staff since Cho came on board. Last year, the Los Angeles Business Journal named it one of the Best Places to Work in greater LA.
To turn around the business, Cho had to fix five fundamental analytics flaws at Evite. He believes these are flaws that any CEO should also take care to avoid.
Flaw No. 1: Underestimating the Voice of the Customer
Businesses often undervalue customer input to their detriment, Cho said. In his first two years on the job, he focused on solving the list of customer complaints coming in through the ‘support’ part of the website. Users pointed out basic issues such as, “My app doesn’t work,” or “I have an Android phone but you don’t have an Android app,” he said. In total, Evite fixed 130 small and big issues over two years. “That actually stabilized user growth.”
As the user experience improved, Evite’s net promoter score rose. This is a metric ranging from -100 to 100 that scores how willing a customer is to recommend a company’s product or service to others. Cho said that even in the doldrums, the net promoter score of Evite’s core, free service was at 65, which was “kind of insane,” due to a loyal, female-centric base. “This was actually one of the reasons we did not take the MySpace route.”
But that score did not include other folks Evite surveyed who weren’t regulars. They gave the service “horrible scores,” Cho said. After fixing the problems and redesigning the site, Evite’s score is now between 75 and 80. The lesson here is “don’t ignore the power of what seems like super-simple input coming from your customers,” he said. “It’s probably the most valuable thing.”
Flaw No. 2: Assuming ‘Math’ Will Overcome Bureaucracy
Good ideas can bubble up in a firm, but for management to act on them, the concepts must align with business objectives. Failure to recognize this reality can stall an idea and discourage employees. “I’ve seen incredibly intelligent and frustrated people whose capabilities are not being used to the maximum effect in the organization because they actually don’t understand the broad bureaucracy in which business operates,” Cho said.
Cho explained that a business has three types of priorities: strategic, organizational and operational. Strategic priorities are those that address business growth over several years; organizational priorities are those that change a company’s capabilities; and operational priorities are immediate concerns that are preventing the business from serving customers well or functioning efficiently. When ideas come up that are aligned to these priorities, “they get done very quickly,” Cho said.
For example, one of the biggest “annoyances” at Evite was dealing with spammers that use its platform to reach people, Cho said. “They send emails that say you’ve been invited to this party with free food. Open it up and you find a link to Viagra or something.” Booting them out of the system was a big headache especially since there were tens of thousands of them. “Our customer support folks were going crazy playing ‘whack-a-mole,’” he said. The engineering team, which was excited about machine learning, proposed using these algorithms to solve the problem.
Cho immediately greenlit the idea. “That was an active operational priority. [The engineering team] had a solution that was much more scalable than what we’ve got, so we gave them the resources and the time to do that,” he said. The same team would have failed if they came forward with a random machine learning idea that conceptually might make sense but didn’t fit Evite’s objectives. “Getting your idea through the system is really about understanding the system and aligning yourselves to it.”
“Our purpose as a business is much broader than being the world’s biggest invitation service. It’s connecting people together face-to-face.”
Flaw No. 3: Optimizing Components vs. Systems
One of the hardest tasks for Cho when managing analysts or business folks in general was training them to look at the bigger picture, not just fixate on fine-tuning operational components. Businesses can get myopic when sometimes what they need was to enact a broader, systemic change. For example, if numbers were the pure consideration, one would get rid of low margin products in a store even if it reduces variety. But looking at the situation systemically to see what brings the largest net overall customer value, “you come to a different answer,” he said.
Going back to Evite’s experience with spammers flooding the site, he said the company saw a 131% rise in invitation gallery users — but they were useless and even served to dampen activities of real users who were creating invitations. “All of that noise going through the system, but it basically did nothing,” he said. “It was all spam.” So, Evite took steps to address it even though it meant taking a short-term hit. The lesson: Look at the totality of the business.
Flaw No. 4: Mistaking Local Maxima for Global Maxima
Global maxima is the highest point one can get to; local maxima is the peak of an area, but it is not the highest overall. Businesses don’t understand the two concepts “so they plug away and incrementally optimize,” Cho said. “They don’t ask themselves, ‘Are you sitting on the right platform?’” And if they do see they’re on the wrong platform, they are nonetheless unable to switch to the right platform “because all of the data is telling them it’s a horrible idea,” he added.
Case in point: Evite’s simplistic site. “We knew we needed to expand the user experience from being more than just opening up an invitation and pressing a button to say yes or no,” Cho said. “I knew strategically that the overall user interface paradigm needed to get blown up.” Evite needed a design that could scale to accommodate multiple functions over time. “This was a huge change” for long-time Evite users, who were used to essentially a one-page functionality without tabs, he said.
“There was crazy fear in my organization when I first introduced this,” Cho continued. “They said, ‘Oh, the users are so used to this. This is going to tank all of our conversion rates.… RSVP rates are going to go down.’” Cho responded with, “Yeah, maybe. Probably. It doesn’t matter.” But the jitters continued. “They said, ‘Oh we should test this!’” Cho said. “I said, no, we shouldn’t test it. They looked at me like I was some data terrorist.”
But Cho insisted. “I’ve been grounded in data. Data’s really important. But there’s no test that we can run that’s going to tell me that this is a good idea,” he said. The reason is that the Evite experience has already been micro-optimized for 15 years to be the best possible one-page experience. It had hit its local maxima. “It was clear to me that it’s better to be on the global maxima,” Cho said. “That was a leap we had to take without the data.” Today, Evite’s app has multiple tabs for different functions such as to RSVP, upload photos or interact in a private activity feed within a closed group, similar to social networks.
“If you have earned the privilege as a business leader to control any levers of scale … you have an obligation to leverage that scale and turn it into some social good.”
Flaw No. 5: Focusing on Step vs. Slope Change
When Cho’s management team goes to him recommending a step change, “I as the CEO always push them and say, ‘Are you thinking too small? Is there a slope change opportunity versus a step change opportunity?” He contends that MySpace would have “never come out of its decline arc with a bunch of rational step changes. It literally needed to reinvent the slope of the business.”
This was Evite’s slope change: moving from emailed invitations to ones by text. “Pretty big change in how Evite shows up for a user,” Cho said. “That has literally changed the underlying arc of the business.” He confessed that the idea of being able to send invitations by text actually didn’t come from within the company; it was the most requested feature by users. Sometimes “the low hanging fruit is so low hanging, it’s massive,” Cho said.
Don’t Forget the Fourth Stakeholder
In the end, Cho said, a company should not just focus on serving three stakeholders — customers, shareholders and employees. It should add a fourth stakeholder: society. Evite now lets users donate to a list of charities on the site. “There are a whole bunch of people throwing a lot of parties. They’re getting together, they’re spending money. They’re probably buying gifts they don’t need,” Cho said. “Let’s see if we can turn that into a social cause.”
Cho’s team had questions. “My product team says, ‘That’s great. What’s the ROI?” he recalled. “I said, ‘There’s no ROI.” The team also was worried that adding a donation feature would affect usage because people would feel guilty if they didn’t donate. “That doesn’t matter,” Cho told them. So far, Evite has raised nearly $7 million through its platform.
“My perspective is if you have earned the privilege as a business leader to control any levers of scale — that could be user scale, that could be financial scale — you have an obligation to leverage that scale and turn it into some social good,” Cho said. “Raise money, do some good.”