AVAILABLE DATA TO reaching wrong conclusions about FORMULATE - PowerPoint PPT Presentation

Avery Vise Principal TransAdvise Profile Image 205-358-3612 avise@transadvise.com @AveryVise OPTIMIZING PUBLICLY Or how to avoid AVAILABLE DATA TO reaching wrong conclusions about FORMULATE DECISIONS the trucking industry

Why FMCSA data? • Easily obtained and cheap – even free • Operational insights, such as where a carrier operates, what kind of freight it hauls, etc. • Minimal lag time, especially for pinpointing new carriers • Available for commercial vehicle owners in all states, sort of Optimizing Publicly Available Data to Formulate Decisions 2

FMCSA databases • Census – Basic demographics, updated at least every two years (mostly) • Inspections – Includes truck/trailer brand and VIN • Violations – Violations discovered during inspections • Crashes – DOT-recordable crashes – i.e., fatality, injury requiring medical attention away from scene, or tow-away • SMS metrics – Activity and measures in each BASIC • Licensing & Insurance (L&I) – Useful in confirming for-hire carriers remain in business. (http://li-public.fmcsa.dot.gov) Optimizing Publicly Available Data to Formulate Decisions 3

Motor carrier annual reports • By law (49 USC 14123), DOT must require carriers with at least $3 million in annual revenue to file annual financial and operations reports (Form M). • In practice , DOT doesn’t enforce the mandate. FMCSA’s latest estimate: Only 308 carriers file. Most of those probably are HHG and publicly traded carriers. • 49 CFR 369.10 states that the data generally is publicly available, but FMCSA insists on a FOIA request, which takes months. The data hasn’t been freely available since 2003. Optimizing Publicly Available Data to Formulate Decisions 4

Two versions of FMCSA data Purchase from FMCSA Monthly free download Website https://ai.fmcsa.dot.gov/SMS/Tools/Downloads.aspx https://ask.fmcsa.dot.gov/app/mcmiscatalog/mcmishome PLUS : County, For-hire vs. private, MC number, types of haul, Dun & Bradstreet number, detailed info on type of power unit and whether owned or leased, number of trailers and whether owned or leased, detailed info on DOT number, company name, address, phone, Census types of driver (CDL vs. non-CDL, range of operation), fax, email, vehicle mileage, carrier age, names (but not titles) of company representatives, more. interstate versus intrastate, number of power units and drivers Cost : $23 per month. PLUS : Violations, more granular info on inspections, DOT number, date, location, type of unit (straight GVW/GCW, shipper name. truck, tractor, bus, trailer, etc.), make of unit, Inspections VIN, license state, safety-related info Cost : $70 per calendar year, but timing matters. DOT number, detailed violation data – Vehicle, Violations HOS, driving, etc., with SMS-related info Included in the inspection file extract PLUS : Greater detail regarding circumstances surrounding crash. Basic info related to crash – DOT number, date, Crashes fatalities, injuries, tow-aways, weather, VIN Cost : $36 for a 24-month extract. Measures and inspections in the BASICS for active SMS Metrics motor carriers Not applicable Optimizing Publicly Available Data to Formulate Decisions 5

Commercial data suppliers www.rigdigbi.com www.tdsource.com www.fleetseek.com • Resolve data hygiene issues (or at least most of them) and simplify filtering and searching • Integrate with other data sources/tools (VIN decoding, SIC codes, subscriber databases, D&B, L&I) to varying degrees • Fill privacy-state gaps in vehicle registration data (i.e., Polk) • Might not offer complete historical data • Don’t solve inherent limitations with underlying data Optimizing Publicly Available Data to Formulate Decisions 6

Potential uses of FMCSA data • Market research and analysis • Fleet size by carrier type, cargo, etc. • Productivity/utilization • Age of motor carriers • Trailer-to-tractor ratios • Vehicle maintenance trends • ELD adoption • Ethnicity of ownership • Much more Optimizing Publicly Available Data to Formulate Decisions 7

Equipment utilization OTR trucking companies Annual miles Fleet size per truck 1 to 9 61,844 10 to 49 63,537 50 to 99 69,642 100 to 499 75,124 500 to 999 77,545 1,000+ * 79,257 Total 70,511 *Excludes UPS and FedEx parcel operations Source: FMCSA census data, July 2017 Optimizing Publicly Available Data to Formulate Decisions 8

ELD adoption OTR trucking companies Carriers Trucks* Fleet size 1 to 9 0.7% 1.4% 10 to 49 8.5% 10.9% 32.0% 50 to 99 33.2% 100 to 499 59.8% 65.1% 500 to 999 85.2% 86.3% 1,000+ 94.5% 96.3% Total 2.2% 46.2% *Represents trucks operated by carriers that show indicator of electronic log use. E-logs are not necessarily installed in all trucks. Source: FMCSA inspection and census data, July 2017 Optimizing Publicly Available Data to Formulate Decisions 9

Over-the-road trucking company • Defined as: • Interstate U.S. • Authorized for-hire, excluding private • At least one truck • At least one CDL driver • At least one driver operating beyond 100 miles • Generally better coverage in FMCSA data than the universe of USDOT-registered carriers Optimizing Publicly Available Data to Formulate Decisions 10

Potential uses of FMCSA data • Market research and analysis • Equipment ownership insights • Make and type of power unit and make of trailer in raw data • VINs can be decoded to provide greater detail Optimizing Publicly Available Data to Formulate Decisions 11

Potential uses of FMCSA data • Market research and analysis • Equipment ownership insights • Prospecting/lead generation • Fleet size • Geography (Domicile and operational footprint) • Type of haul • Brand affinity • Maintenance violations • Crash types Optimizing Publicly Available Data to Formulate Decisions 12

Potential uses of FMCSA data • Market research and analysis • Equipment ownership insights • Prospecting/lead generation • Customer relationship management • Appending FMCSA data could provide useful insights • Examples: • Truck dealer could identify customers who acquire competitors’ trucks • Creditors and investors could monitor carriers’ shipper base for potential revenue and freight volume changes Optimizing Publicly Available Data to Formulate Decisions 13

Optimizing Publicly Available Data to Formulate Decisions 14

Shippers and carriers Selected carriers Selected shippers • Andrus Transportation Services Inc. • Bridgestone • Big G Express Inc. • Goodman Manufacturing • DPI Dedicated Logistics Inc. • Jack Daniels • Interstate Distributor Co. • Mars Petcare • ITS Logistics LLC • Nestlé Waters North America • Mann Logistics LLC • Owens Brockway Glass Container • Mile Hi Specialty Foods Inc. • Restoration Hardware • Swift Transportation Co. • Starbucks Optimizing Publicly Available Data to Formulate Decisions 15

Potential uses of FMCSA data • Market research and analysis • Equipment ownership insights • Prospecting/lead generation • Customer relationship management • Safety analysis Optimizing Publicly Available Data to Formulate Decisions 16

The reality of FMCSA data • Census • Requires significant scrubbing and filtering Optimizing Publicly Available Data to Formulate Decisions 17

FMCSA census data What is an active carrier? Active July 2017 U.S. FMCSA definition: • In the past 3 years, carrier has had: Entities 1,480,482 • An inspection • + A crash 30,612,502 • A compliance review • 6,450,624 A safety audit • An MCS-150 update + + 37,063,126 • Vehicle registration activity • A Unified Carrier Registration (UCR) 18,337,667 system payment; 12,229,835 OR 16,034,543 • Carrier holds current operating authority indicated in the L&I database. 9,270,659 Optimizing Publicly Available Data to Formulate Decisions 18

Bad data in the FMCSA census Disclaimer on FMCSA website: “FMCSA uses data filters, tailored scripts, and quality assurance checks when generating official statistics from the raw data it receives from motor carriers. The data you are receiving/purchasing are raw data, and have not been filtered or checked for accuracy.” Optimizing Publicly Available Data to Formulate Decisions 19

Bad data in the FMCSA census Optimizing Publicly Available Data to Formulate Decisions 20

Bad data in the FMCSA census Optimizing Publicly Available Data to Formulate Decisions 21