A tale of two UBI programs, as told over Twitter

October 7, 2016 / 8 min read

A large volume of surveys, analyses, and speculation over recent years has focused on understanding how consumers feel about usage-based auto insurance (UBI). Ironically much of this activity fails to consider what consumers themselves are actually saying in publicly available forums such as social media.

Conventional wisdom rightfully suggests that discount incentives and access to value-added services may help ingratiate UBI to policyholders. But before any such benefits to consumers change hands, insurers must gain (remote) access to policyholders’ vehicles, measure driving habits, and form actionable insights. How effectively (or ineffectively) insurers implement these steps may be a major differentiator between the most and least successful programs. This is a reality that is profoundly echoed by hundreds of social media posts on the topic of UBI that we analyzed.

twitter UBI hand web

Looking at the Twitter data

In order to understand the principles that drive effective implementations, we analyzed a sample of microblogs ‘tweeted’ in 2015 related to two well-known UBI programs in the United States. One program is offered by a larger insurer who was an early mover and the other by a relatively new insurer who operates in a handful of jurisdictions.

We archived tweets related to these programs using the keyword archiving capability offered by Hootsuite. We then employed sampling so we could manually inspect and classify a manageable number of tweets. Content was limited to “tweeters” who clearly participated in the insurers’ UBI programs and was then classified into categories such as privacy, advertising, technology, and pricing.¹ Such classifications may serve as the basis for “topic models,” which algorithmically classify larger numbers of tweets in real time. Analyzing similarities, differences, and themes in how the two programs are perceived provides useful insights into how insurers may master the art of UBI. Here are five lessons we took away from the analysis:

Provide a seamless technological experience

Approximately two out of every five tweets about the larger insurer’s program related to the technological experience—with a vast majority bemoaning it. Specific complaints included the inconvenience of installing the insurer’s “dongle” device in a vehicle’s OBD-II port; incompatibility with certain vehicle models; alleged damage to vehicles’ batteries, alternators, and computers; and confusing or frustrating audio messages. By contrast, nearly half of the technological feedback about the smaller insurer’s program, which features both a device and smartphone app, was positive.

Despite occasional complaints resembling the tweets about the larger insurer, many more tweets about the smaller insurer lauded information provided through the app, such as vehicle health reports or alerts about potentially “ticket-able” situations. Technological experience may serve as a UBI program’s “first impression,” with policyholders, so the easier the installation (if installation is required), the higher quality the technology used, and the more engaging and insightful the monitoring process, the more likely it is that the consumer’s overall impression will be positive in the end.

Provide feedback that makes sense

The larger insurer’s program measures risk factors such as mileage, times of day of operation, and changes in velocity (acceleration and braking), while the smaller insurer markets a “pay-per-mile” approach that is invariant with respect to how riskily the vehicle is operated. While the larger insurer’s program may allow for more driving time as long as it is done responsibly, many of the technological complaints noted above suggest that that insurer’s definition of “risky driving” warrants further consideration.

Various tweets point to hiccups in the larger insurer’s methodology. For example, in their program, downhill driving might register as “harsh” acceleration. Shifting gears or engaging a parking brake could compute as “harsh” braking, while driving in slippery weather or dense traffic could be flagged as excessive occurrences of acceleration and braking. And, most commonly, the device could potentially encourage drivers to ignore traffic laws by flagging stops at “quick lights” as “harsh” braking.

By focusing primarily on mileage and communicating ancillary information that may be cost-saving outside the insurance transaction (e.g., vehicle health), the smaller insurer appears to have largely insulated itself from such criticism. This is not to say risky driving cannot be assessed effectively, but insurers doing so may wish to employ geographic information system (GIS) data to contextualize incidents. For instance, they may consider proximity to lighted intersections to help determine whether or not braking is appropriate.

Under-promise and over-deliver

The worst-kept secret in UBI may be that policyholders enroll seeking discounts. Approximately one out of five tweets about both insurers’ programs related to policy pricing. However, up to a quarter of the larger insurer’s pricing-related tweets were negative.

For example, consumers mocked or questioned the diminutive size of their discount relative to advertised possibilities, or in some cases remarked on a premium increase (likely unrelated to UBI, and perhaps caused by some offsetting factor such as an accident surcharge). By contrast, such complaints were relatively rare in the newer insurer’s Twitter feed. Many policyholder cited savings of 40, 50, or even 60 percent relative to what they were previously paying.

The smaller insurer’s record of over delivery was not limited to discounts. Our analysis also revealed cases where the company’s staff engaged constructively with participants; for example, helping to resolve installation or data interpretation issues, thereby “over delivering” on service.

Both programs’ streams contained tweets from consumers who wished to use data for non-insurance purposes, such as for speeding ticket challenges or to defend IRS audits, providing another opportunity to save money. These anecdotes prove that by establishing an ongoing connection in the vehicle, UBI provides the insurer with a plethora of opportunities to exceed consumers’ expectations.

Embrace differences between policyholders

UBI means different things to different people, so a “one-size-fits-all” strategy may miss the mark. Many participants in both programs tweeted with tongue in cheek about their radio sing-alongs and rearview mirror hair grooming sessions being subject to UBI monitoring, but policyholders’ reasons for allowing insurers into their vehicles were as varied as their behind-the-wheel activities.

One participant tweeted: “You can Big Brother me all you want if it means cheaper premiums.” This sentiment was typical of consumers in both feeds, who identified themselves as garaging in traditionally high insurance cost areas such as Detroit, Chicago, or the Bay Area, markets the smaller insurer’s program catered to.

By contrast, a frequent source of relief for parents (and a source of good-natured lament for their teenaged children) in the larger insurer’s stream was the ability to use feedback provided by the program to help ensure the safety of newer drivers. The larger insurer’s program may be better positioned to connect with the parental segment because it focuses on behaviors such as maneuvering and times of day of vehicle operation.

Now consider if we provided one insurer’s program to the other’s “core constituents,” and vice versa. Complex maneuvers and frequent braking are virtually inevitable in metropolitan areas with congested traffic conditions, so the larger insurer’s program, which rates policyholders on such behaviors, may not appeal to the smaller insurer’s target urban markets.

And parents who enroll their newer drivers’ vehicles in UBI may be less interested in low mileage incentives or ticket-preventing street-sweeping alerts, which are provided under the smaller insurer’s program, and more interested in knowing that their loved ones aren’t drag-racing or driving during late night hours—the types of behaviors that are measured by the larger insurer’s program.

The two programs’ target markets are not necessarily mutually exclusive, but both insurers do seem to have connected with a certain niche. These anecdotes demonstrate the value of a clear and well-articulated UBI program in terms of product design, marketing, and distribution.

Get in the game

Approximately one quarter of tweets about the larger insurer’s program and one half of the tweets about the smaller insurer’s program were positive in nature. Based on this, one may be tempted to conclude that the latter’s program is “better,” but social media data tends to skew cheeky, and the larger insurer’s more recognizable brand presented a larger target for criticism.

Although our Twitter analysis revealed areas where each insurer may be experiencing greater success, the major victory for both is that they’ve created engaging games of one-upmanship which policyholders are spiritedly discussing online. Many consumers tweeted positively about their ability to control their outcomes under both insurers’ programs, with one consumer claiming to be “the first person to tweak [smaller insurer’s] policy… to better arbitrage my per-mile payout,” and another claiming “[the larger insurer] is about to get punked when I drive like a grandma these next months … with this contraption.”

While some policyholders may eventually revert to riskier habits after UBI, more drivers may find safer tendencies become ingrained, resulting in cost savings and reduced heartache for insurer and insured alike. This truth is not lost on policyholders who, for instance, tweeted “thanks to [larger insurer’s program] I was able to prove who really is a better driver, and of course it’s me … brother and dad [stink],” or “I’ve mastered [larger insurer’s] hard brake concept.” Insurers developing UBI programs that captivate and serve in this way are leading the race for policyholders’ hearts, wallets, and tweets.

Implications for insurers

Social media data has significant value in capturing the consumer zeitgeist. Financial institutions are beginning to consider social media as a predictive data source for assessing an individual’s creditworthiness. This information may be useful in resolving accessibility issues and improving transactional costs related to consumers with limited financial histories—similar to benefits typically associated with UBI.

Today’s consumers have long tapped into the social media grapevine to evaluate insurers and to make their preferences—and grievances—known. Insurers can in turn use social media to evaluate policyholders, better understand their behaviors, and listen to their sentiments.

To be sure, the voice of the consumer can be heard loud and clear across social media platforms. Those insurers who carefully listen and respond will be the ones who successfully connect with them.