Friday, January 15, 2021

Sometimes Access Speed Does Not Differentiate in a Positive Way

Though U.S. industries get varying scores on the American Customer Satisfaction Index, there sometimes is more divergence within each category than between categories. Consider internet access service. 


In the ISP space, for example, Frontier Communications gets unusually low scores, compared to Verizon’s fiber to home service, likely because Frontier connections use slow digital subscriber line platforms, much of the time. 


It is harder to explain relatively poorer satisfaction with Altice or Mediacom services, as both seem to offer high broadband speeds. Presumably customer operations or prices are the issue there. 


Consider also that CenturyLink (now Lumen Technologies) scores the same as Charter Communications, though Lumen access speeds are decidedly lower than Charter’s speeds. Where Charter customers get an average of 95 Mbps, Lumen customers get about 35 Mbps. 


Access speed alone does not seem to explain very much. Something else in the experience seems to be operating. 

source: ACSI   


Customer satisfaction ratings for the linear video subscription business and internet service provider businesses historically have ranked at the very bottom of all industries measured by the American Customer Satisfaction Index. Mobile service has ranked higher in recent years, indeed scoring the highest of all telecommunications services. 


Since 1995, fixed network phone service has dropped significantly from fairly high levels of satisfaction. That might require some interpretation. Most U.S. consumers do not buy the product anymore, so the remaining customers seem to be showing modest satisfaction. Presumably the non-buyers consider the product to be unsatisfying. 


Video streaming services rank higher than linear video services, which might make sense given the growth of streaming services and the slow decline of linear services.


No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...