Mediacom: ACSI Samples Distort

Mediacom Communications has been trying hard to put out the PR fires that have sprung up from its last-place showing in the widely reported American Customer Satisfaction Index, adding that the sample size used to determine the small market operator’s ACSI rating unfairly distorts its true customer service record.

Mediacom finished in last place in the ACSI survey, with a score of 57 out of 100 in Internet services and 54 out of 100 for subscription TV services, the lowest among the 43 industries that ACSI tracks. The TV  distinction caused many news outlets (not this one) to tout Mediacom as “Most Hated Company in America.”

ACSI stands behind its survey and said that although Mediacom did have a smaller sample size for the study, it is nonetheless accurate.

The unwanted designation comes at perhaps the worst time for the small market operator. It recently celebrated its 20th Anniversary, kicking off that celebration with a $1 billion capital investment plan that will boost residential data speeds in its markets to 1-gigabit per second, expand its business services capabilities, extend fiber deeper into its network and deploy community WiFi access points throughout high traffic commercial and public areas across its footprint.  In addition, like other larger and smaller cable operators, Mediacom has made strides in customer service, initiating nights and weekends service calls and 30-minute windows for appointments.

Mediacom also believes it is being unfairly compared to much larger operators and that because of its size, the number of respondents to the ACSI survey in its service territory was small. According to Mediacom senior vice president, government & public relations Tom Larsen, of the 12,710 people that participated in the survey, only 95 were Mediacom customers. He said the usual minimum sample size for ACSI is between 100 and 250 respondents. What’s more, the margin for error for the Mediacom portion is plus or minus 6 points, twice the 3-point error margin for the rest of the companies in the study.

“A 3 point versus 6 point margin of error is significant in this type of survey when just a few points separate so many of the companies,” Larsen said in an e-mail message. 

Larsen is right. The top company in the subscription TV survey (Verizon) scored a 70. An extra 6 points would have placed Mediacom even with Charter, which had a score of 60.

ACSI director of research Dr. Forrest Morgeson said the researcher does like to measure a minimum of between 200 and 250 responses for each company, but obviously sometimes that isn’t possible.

“For companies like Comcast and Charter, we can do 200-to-250 interviews in a day,” Morgeson said. “But as you go down the ladder, it gets harder to do data collection and we will allow smaller sample sizes. That’s what happened to them [Mediacom]. We weren’t trying to short change them. We wanted to include them.”

Morgeson added that while there is a larger confidence interval with smaller sample sizes – about 6 or 7 points he said – Mediacom’s claims that the study couldn’t be trusted is “wrong and self-serving.”

“We’re very confident in the results,” Morgeson said, adding that ACSI also looked at past studies from other researchers and found that Mediacom’s results were strikingly similar.

And there was one other small market operator that had a similarly sized sample that fared better – Suddenlink Communications, which Morgeson said had a sample size of just under 100. Mediacom, he said, had a sample size of 95 for both ISP services and subscription TV, or about 50 for each segment.

Suddenlink had a subscription TV ACSI score of 62 (8th place, tied with Comcast)and an ISP score of 61 (10th place, ahead of Comcast)  which put them around the average score of 65.