Belgian OOH survey 2017- Ready for the future - FEPE · Step 5 Step 6 Travel data 3 steps Traffic volumes 3 steps. ... Solution Yes No ... •Up to date traffic data based on measured

Belgian OOHsurvey 2017-

Ready for the future

What does the market need?

• Reliable OOH-survey

• Covers the whole country

• Survey with up to date data

• High granularity

• Compatible with DOOH

• Integration of visibility insights

• Future-proof

A future-proof study

• Mix survey data with big travel data

• Modular approach allows integration new data sources

• Integrates all available visibility insights

• Allows real contact based NEW currency

• Will support new business models

Visibility measurement

Traffic measurement

Partners

Why Be-Mobile

- Performance of Out-Of-Home advertising dependshighly on quantity & quality of eyeballs

- Classic approach: surveys- Very limited sample in time and population- Complexity of survey

- Collecting more data seems necessary, but- Sensor networks are very expensive- Existing big data offerings (eg. Telco data)

are not sufficiently detailed

- Solution: Be-Mobile Mobility Data

Be-Mobile

mobility

database

Floating car data

• GPS data

• Smartphone data

• Telco data

Road sensors

• Cameras

• Radars

• Loops

• Parking sensors

Crowd sourced

• Drivers

• Social media

• Police

• Emergency vehicles

Other data

• Public transport

• Car & bike sharing

• Parking, fuel

• Toll

3.000.000.000 data

records/day6.000.000 vehicles

monitored

The most versatile mobility data aggregator

How can we transform the Big Data into real traffic?

Activity based modelling

Virtual Population (VP)

Database

Survey

Data

Week travel diaries

for VP

Validation with

passive traffic data

Map trips

on road network

Calibration

of travel data

Step 1

Step 2

Step 3

Step 4

Step 5

Step 6

Travel dataTravel dataTravel dataTravel data

3 steps3 steps3 steps3 steps

Traffic volumesTraffic volumesTraffic volumesTraffic volumes

3 steps3 steps3 steps3 steps

Step 1 : Survey data

• Integration of 3 sources

• CIM OOH study

• BELDAM (Belgian government)

• OVG (Flemish government)

• Missing data estimated

Alignment of socio-demographics

• Output = 45.000 respondents

Step 2 : Virtual Population dbase

• Create dbase 9,6 mio Belgians (12+)

• Basic SD = federal census data 2011

• Estimate missing SD’s

• Using survey data to extrapolate SD-patterns

on clusters of statistical zones of similar composition

Step 3 : Week travel diaries for VP

Create 180 million trips for 9,6M Belgians, for 1 week

Trip choices

Frequency by motive How many trips does an individual make

in a week to his work/school, to…?

Time-of-week + sequence And on which day and day part?

Destination What is the destination of his trip?

Transport mode What is the transport mode for each trip?

All values predicted following a model based on the survey data(using gender, age, occupation, family size, car ownership, pop. density, trip motive / distance / travel time)

Activity based models validated on survey data

Step 3 : Week travel diaries for VP

E.g. School trips per week for all 12-24 year old in survey and in model

Model Survey

Travel dataTravel dataTravel dataTravel data

3 steps3 steps3 steps3 steps

Virtual Population

Database

Survey

Data

Week travel diaries

for Virtual Population

Validation with

passive traffic data

Map trips

on road network

Calibration

of travel data

Traffic volumesTraffic volumesTraffic volumesTraffic volumes

3 steps3 steps3 steps3 steps

Step 1

Step 2

Step 3

Step 4

Step 5

Step 6

Step 4 : Map trips on roads

All 180 million trips get the most probable route at a specific moment

taking into account duration and distance

Route at 08:00 Route at 17:00

Step 5 : Calibration / Validation with passive travel data

• CAR volumes (total) => FOD Mobility & DIV

• CAR volumes (per segment) => 700 Loop detectors & 3% GPS

• Total pedestrian, bicycle, PT => Surveys

• Public transport by station => Traffic counts stations

Travels

from diaries

Estimated

Traffic Volumes

Trip mapping

Validation

Solution

Yes

No

Ca

libra

tion

• Increase amount of short trips

• Increase/reduce mean trip lengths

• Shift between time-periods

• Shift between modes

Step 6 : Calibration of travel data

• Up to date traffic data based on measured big data

• Real time speed

• Audience measurement will deliver even more granularity in the future with audience measurement for each hour of the day

• Breakthrough for DOOH

• Projected audience measurement by hour• Different VAI for each hour every day

Advantages

• New commercial model for DOOH

• Dynamic but transparent pricing system for DOOH

• Data needed for automated planning/buying

Opportunities

• Hybrid future proof modular survey

• Audience measurement based on real accountable data by independent third parties

• Fundamentals for a new commercial model in DOOH

Conclusion