As part of an FCC effort to improve their broadband data collection, the company last year hired UK firm SamKnows to provide a better glimpse at the real speeds consumers were seeing. SamKnows gave 9,000 volunteers home routers with custom firmware designed to monitor daily connection performance, and the FCC today released their first report based on that data. The report examines the performance of thirteen major U.S. ISPs, and by and large shows that most ISPs deliver at least 80% of the speeds advertised. However, some ISPs perform better than others in this regard, and several ISPs aren't delivering the speeds consumers are paying plenty for -- particularly at peak hours.

The report (pdf) notes that ISPs helped their cause immensely by doing a better job accurately advertising their speeds in the first place, many eliminating the dreaded "up to" marketing metric. The study also highlights the differences in delivered speeds by technology, with the data showing that DSL delivered 88% of advertised downstream speed during peak, compared to 93% for cable broadband and 113% for fiber to the home services. Upload speeds during peak hours were 95% of the advertised speeds for DSL, 108% for cable and 112% for fiber.

When broken down by ISP, the report is particularly unkind to Cablevision -- who was found to deliver just 50% of actual downstream sustained speeds advertised during peak usage hours. Frontier also makes an uninspired showing, delivering just 67% of advertised speeds during peak hours. Windstream, Mediacom, AT&T and Qwest also struggled to meet advertised speeds, though their consistency is significantly better than Cablevision.

Not too surprisingly, ISPs that have been engaged in significant upgrade efforts fare much better. Comcast, who has been busily upgrading their network to DOCSIS 3.0 technology, delivers 144% of the sustained downstream speeds advertised. Verizon's FiOS FTTH customers see 104% of peak advertised speeds.

Previous reports by the FCC had claimed that consumers were seeing as low as 50% of the speeds they're paying for, though as we said at the time -- most people in the industry thought that claim to be somewhat inflated. Still, the FCC's new real-world data clearly highlights there's a gap when it comes to following through to advertised speeds, particularly for phone companies nursing aging, distance-constrained DSL technologies. Obviously the study doesn't address the fact U.S. consumers pay more for these services due to limited competition (or the fact the FCC by and large doesn't want to address this fact), but it remains promising that the FCC has at least stopped using completely imaginary data to inform policy decisions.