What is IP matching?
IP address matching is leveraged to recruit respondents when the majority of impressions from a given campaign, site, or placement are delivered on media properties where a device-level identifier (AAID, IDFA etc) cannot be shared (in cookieless, and other id-less environments).
Common examples include properties and placements that are delivered to CTV devices, or certain podcast ad servers that cannot pass the mobile ad ids in real-time. Examples include, but are not limited to:
Connected TV devices (Roku, Chromecast)
Linear TV (Samba, iSpot Data) - crosswalk is generated for each data partner to match linear TV IP addresses to respondents
Mobile in-app environments where an individual’s IDFA or AAID is inaccessible
Wearables
An Internet Protocol address (IP address) is a numerical label assigned to each device (e.g., computer, printer) participating in a computer network (such as your home network).
The basic idea is to find consumers exposed to a campaign on a connected device by comparing the IP address when they were exposed to the ad against IP address when they come in to do a survey. When the IP addresses match, there is a high degree of probability that these are the same devices.
Accuracy of IP matching
It is difficult to estimate the error rate by using IP matching because IP addresses reset on a regular basis. Therefore, the smaller the lag between the IP capture at exposure and the panelist match, the higher the likelihood that this is the same device. If you are comparing IP addresses that were collected 30 days apart, the error rate is extremely high (Cint estimate 80%+) and we would not recommend an IP match.
The benefit of the Cint system is that we use real-time IP data for the panelist (not historical). Because of this, if the respondent was exposed yesterday and the survey is taking place today, Cint selects the respondent based on a match within a 48 hour lookback window, increasing the accuracy of the match. This is different from traditional panel approaches that rely on historical data that may be weeks or months old to select the respondent.
Lucid Measurement uses metadata, such as device, as well as baseline and qualification questions to assess a respondents’ media consumption behaviors and ensure the respondent who completes the survey was likely the same person in the household who was exposed to the ad.
See baseline questions, linear TV baseline questions, and CTV streaming qualification questions for more information.
Lucid Measurement process for IP matching
Media tracking IP collection
IP Addresses are collected by Cint’s tracking tag on each impression event. This information is stored alongside campaign impression metadata for 366 days.
IP Addresses are also collected via S2S data processes, where Lucid receives a direct feed of exposure information from a partner. Examples include iSpot TV and Samba TV, where IP is collected by smart TV operating systems.
Cint sometimes receives ad hoc IP targeting lists from clients to use for custom projects. This practice has diminished as companies have begun cracking down on sharing IP address information outside of secure, S2S transfer processes for privacy reasons.
Lucid Marketplace IP collection
IP address is collected on every respondent who enters the Lucid Marketplace. Every respondent who enters Marketplace is required to have a valid IP for security reasons.
Marketplace asks all respondents for consent to leverage their IP address for survey recruitment.
Recruitment process
Cint collects an IP address via media tracking methods (Pixel or S2S).
When a respondent enters Marketplace, Cint looks up the 2-day exposure history for that respondent’s IP along with the sites, placements of exposure.
Surveys that have IP matching enabled may qualify for an IP matched opportunity for exposed.
If IP matching is enabled, it will also prevent a respondent from receiving a control survey if there was a recorded exposure to that IP in the past 90 days.