METHODS AND APPARATUS TO GENERATE AUDIENCE METRICS USING THIRD-PARTY PRIVACY-PROTECTED CLOUD ENVIRONMENTS
An example system disclosed herein includes programmable circuitry to identify audience measurement panelists associated with database proprietor accounts, when a sign- out percentage of the audience measurement panelists is zero, determine a first probability that a second server of a database proprietor associated a first demographic with a logged impression when the logged impression corresponds to a second demographic, when the sign-out percentage of the audience measurement panelists is one hundred, determine a second probability that the second server of the database proprietor associated the first demographic with the logged impression when the logged impression corresponds to the second demographic, determine an adjustment factor based on a weighted combination of the first and second probabilities, and apply the adjustment factor to database proprietor impressions data to reduce a computer-generated error.
1 . A system comprising:
memory;
programmable circuitry; and
instructions to cause the programmable circuitry to:
identify, based on merged data stored in computer memory of a privacy-protected cloud environment, audience measurement panelists associated with database proprietor accounts, the merged data including panel data collected by a first server of an audience measurement entity and database proprietor impressions data collected by a second server of a database proprietor based on network communications from client devices, the database proprietor impressions data including a computer-generated error based on the network communications from the client devices;
when a sign-out percentage of the audience measurement panelists is zero, determine a first probability that the second server of the database proprietor associated a first demographic with a logged impression when the logged impression corresponds to a second demographic;
when the sign-out percentage of the audience measurement panelists is one hundred, determine a second probability that the second server of the database proprietor associated the first demographic with the logged impression when the logged impression corresponds to the second demographic;
determine an adjustment factor based on a weighted combination of the first and second probabilities; and
apply the adjustment factor to the database proprietor impressions data to reduce the computer-generated error.
2 . The system of claim 1 , wherein the instructions are to cause the programmable circuitry to:
determine a first weighted probability by weighting the first probability by a sign-in percentage of the audience measurement panelists;
determine a second weighted probability by weighting the second probability by the sign-in percentage subtracted from one hundred percent; and
determine the weighted combination based on an aggregation of the first and second weighted probabilities.
3 . The system of claim 2 , wherein the instructions are to cause the programmable circuitry to determine the sign-in percentage by dividing (a) a first count of first impressions represented in the database proprietor impressions data, the first count corresponding to first media accesses by the audience measurement panelists while signed into the database proprietor accounts, by (b) a second count of second impressions represented in the panel data, the second count corresponding to the first media accesses and second media accesses, the second media accesses by the audience measurement panelists while signed out of the database proprietor accounts.
4 . The system of claim 1 , wherein the instructions are to cause the programmable circuitry to determine the first and second probabilities based on survey data.
5 . The system of claim 1 , wherein the instructions are to cause the programmable circuitry to:
determine a first adjustment factor for first ones of the audience measurement panelists corresponding to a first household and a second adjustment factor for second ones of the audience measurement panelists corresponding to a second household; and
generate an aggregate adjustment factor based on the first and second adjustment factors.
6 . The system of claim 1 , wherein the computer-generated error results from non-coverage of one or more impressions represented in the database proprietor impressions data, the one or more impressions corresponding to a user for which demographics are not identifiable by the second server of the database proprietor.
7 . The system of claim 1 , wherein the computer-generated error results from misattribution of one or more impressions represented in the database proprietor impressions data, the one or more impressions associated with a first user when the one or more impressions correspond to a second user different from the first user.
8 . A non-transitory computer readable medium comprising instructions that, when executed, cause programmable circuitry to at least:
identify, based on merged data stored in computer memory of a privacy-protected cloud environment, audience measurement panelists associated with database proprietor accounts, the merged data including panel data collected by a first server of an audience measurement entity and database proprietor impressions data collected by a second server of a database proprietor based on network communications from client devices, the database proprietor impressions data including a computer-generated error based on the network communications from the client devices;
when a sign-out percentage of the audience measurement panelists is zero, determine a first probability that the second server of the database proprietor associated a first demographic with a logged impression when the logged impression corresponds to a second demographic;
when the sign-out percentage of the audience measurement panelists is one hundred, determine a second probability that the second server of the database proprietor associated the first demographic with the logged impression when the logged impression corresponds to the second demographic;
determine an adjustment factor based on a weighted combination of the first and second probabilities; and
apply the adjustment factor to the database proprietor impressions data to reduce the computer-generated error.
9 . The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed, are to cause the programmable circuitry to:
determine a first weighted probability by weighting the first probability by a sign-in percentage of the audience measurement panelists;
determine a second weighted probability by weighting the second probability by the sign-in percentage subtracted from one hundred percent; and
determine the weighted combination based on an aggregation of the first and second weighted probabilities.
10 . The non-transitory computer readable medium of claim 9 , wherein the instructions, when executed, are to cause the programmable circuitry to determine the sign-in percentage by dividing (a) a first count of first impressions represented in the database proprietor impressions data, the first count corresponding to first media accesses by the audience measurement panelists while signed into the database proprietor accounts, by (b) a second count of second impressions represented in the panel data, the second count corresponding to the first media accesses and second media accesses, the second media accesses by the audience measurement panelists while signed out of the database proprietor accounts.
11 . The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed, are to cause the programmable circuitry to determine the first and second probabilities based on survey data.
12 . The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed, are to cause the programmable circuitry to:
determine a first adjustment factor for first ones of the audience measurement panelists corresponding to a first household and a second adjustment factor for second ones of the audience measurement panelists corresponding to a second household; and
generate an aggregate adjustment factor based on the first and second adjustment factors.
13 . The non-transitory computer readable medium of claim 8 , wherein the computer-generated error results from non-coverage of one or more impressions represented in the database proprietor impressions data, the one or more impressions corresponding to a user for which demographics are not identifiable by the second server of the database proprietor.
14 . The non-transitory computer readable medium of claim 8 , wherein the computer-generated error results from misattribution of one or more impressions represented in the database proprietor impressions data, the one or more impressions associated with a first user when the one or more impressions correspond to a second user different from the first user.
15 . A method comprising:
identifying, based on merged data stored in computer memory of a privacy-protected cloud environment, audience measurement panelists associated with database proprietor accounts, the merged data including panel data collected by a first server of an audience measurement entity and database proprietor impressions data collected by a second server of a database proprietor based on network communications from client devices, the database proprietor impressions data including a computer-generated error based on the network communications from the client devices;
when a sign-out percentage of the audience measurement panelists is zero, determining a first probability that the second server of the database proprietor associated a first demographic with a logged impression when the logged impression corresponds to a second demographic;
when the sign-out percentage of the audience measurement panelists is one hundred, determining a second probability that the second server of the database proprietor associated the first demographic with the logged impression when the logged impression corresponds to the second demographic;
determining an adjustment factor based on a weighted combination of the first and second probabilities; and
applying the adjustment factor to the database proprietor impressions data to reduce the computer-generated error.
16 . The method of claim 15 , further including:
determining a first weighted probability by weighting the first probability by a sign-in percentage of the audience measurement panelists;
determining a second weighted probability by weighting the second probability by the sign-in percentage subtracted from one hundred percent; and
determining the weighted combination based on an aggregation of the first and second weighted probabilities.
17 . The method of claim 16 , further including determining the sign-in percentage by dividing (a) a first count of first impressions represented in the database proprietor impressions data, the first count corresponding to first media accesses by the audience measurement panelists while signed into the database proprietor accounts, by (b) a second count of second impressions represented in the panel data, the second count corresponding to the first media accesses and second media accesses, the second media accesses by the audience measurement panelists while signed out of the database proprietor accounts.
18 . The method of claim 15 , further including determining the first and second probabilities based on survey data.
19 . The method of claim 15 , further including:
determining a first adjustment factor for first ones of the audience measurement panelists corresponding to a first household and a second adjustment factor for second ones of the audience measurement panelists corresponding to a second household; and
generating an aggregate adjustment factor based on the first and second adjustment factors.
20 . The method of claim 15 , wherein the computer-generated error results from non-coverage of one or more impressions represented in the database proprietor impressions data, the one or more impressions corresponding to a user for which demographics are not identifiable by the second server of the database proprietor.
21 . The method of claim 15 , wherein the computer-generated error results from misattribution of one or more impressions represented in the database proprietor impressions data, the one or more impressions associated with a first user when the one or more impressions correspond to a second user different from the first user.