Labeling Examples

Sample hit data hit

Suppose you have the following hit data:

  • The first row contains the labels for each variable.
  • The second row is the name of the variable. If it has an ID label, it contains the assigned namespace in parentheses.
  • Hit data starts in the third row.
Labels
I2
ID-PERSON
DEL-PERSON
ACC-PERSON
I2
ID-DEVICE
DEL-DEVICE
ACC-ALL
I2
DEL-PERSON
ACC-PERSON
I2
DEL-DEVICE
DEL-PERSON
ACC-ALL
I2
ID-DEVICE
DEL-DEVICE
ACC-ALL
Variable Name
(Namespace)
MyProp1
(user)
Visitor ID
(AAID)
MyEvar1
MyEvar2
MyEvar3
(xyz)
Hit Data
Mary
77
A
M
X
Mary
88
B
N
Y
Mary
99
C
O
Z
John
77
D
P
W
John
88
E
N
U
John
44
F
Q
V
John
55
G
R
X
Alice
66
A
N
Z

Sample Access request access

If you submit an access request, you will receive two files that you can return to the data subject. One file is a CSV file containing one row for each hit received for the data subject and a column for each variable with the appropriate access label. The other file is a summary HTML file that lists each variable, followed by all unique values seen for that variable for the data subject and the number of times each unique value was seen.

For our example, the summary file contains the values indicated in the table below. A request may return only a device file, only a person file or one of each. Two summary files are returned only if a person ID is used and expandIds is true.

API Values
Summary
file type
returned
Data in Summary Access File
Namespace/ID
expandIDs
MyProp1
Visitor ID
MyEvar1
MyEvar2
MyEvar3
AAID=77
false
device
not present
77
not present
M, P
X, W
AAID=77
true
device
not present
77
not present
M, P
X, W
user=Mary
false
person
Mary
77, 88, 99
A, B, C
M, N, O
X, Y, Z
user=Mary
true
person
Mary
77, 88, 99
A, B, C
M, N, O
X, Y, Z
device
not present
77, 88
A, B, C
N, P
U, W
user=Mary
AAID=66
true
person
Mary
77, 88, 99
A, B, C
M, N, O
X, Y, Z
device
not present
66, 77, 88
A, B, C
N, P
U, W, Z
xyz=X
false
device
not present
55, 77
not present
M, R
X
xyz=X
true
device
not present
55, 77
not present
M, P, R
W, X

Notice that the setting for expandIDs does not make any difference to the output when a cookie ID is used.

Sample Delete requests delete

With a delete request using the API values in the first row of the table, the hit table will be updated to look something like this:

AAID=77
(expandIDs value does not matter)
MyProp1
AAID
MyEvar1
MyEvar2
MyEvar3
Mary
42
A
Privacy-7398
Privacy-9152
Mary
88
B
N
Y
Mary
99
C
O
Z
John
42
D
Privacy-1866
Privacy-8216
John
88
E
N
U
John
44
F
Q
V
John
55
G
R
X
Alice
66
A
N
Z
NOTE
Only columns on rows containing AAID=77 and a DEL-DEVICE label are impacted.
user=Mary
expandIDs=false
MyProp1
AAID
MyEvar1
MyEvar2
MyEvar3
Privacy-0523
77
Privacy-1866
Privacy-3681
X
Privacy-0523
88
Privacy-2178
Privacy-1975
Y
Privacy-0523
99
Privacy-9045
Privacy-2864
Z
John
77
D
P
W
John
88
E
N
U
John
44
F
Q
V
John
55
G
R
X
Alice
66
A
N
Z
NOTE
Only celcolumnsls on rows containing user=Mary and a DEL-PERSON label are impacted. Also, in practice, the variable containing A_ID would probably be a prop or an eVar. Its replacement value would be a string starting with Privacy-, followed by a random number (GUID), rather than replacing the numeric value with a different, random numeric value.
user=Mary
expandIDs=true
MyProp1
AAID
MyEvar1
MyEvar2
MyEvar3
Privacy-5782
09
Privacy-0859
Privacy-8183
Privacy-9152
Privacy-5782
16
Privacy-6104
Privacy-2911
Privacy-6821
Privacy-5782
83
Privacy-2714
Privacy-0219
Privacy-4395
John
09
D
Privacy-8454
Privacy-8216
John
16
E
Privacy-2911
Privacy-2930
John
44
F
Q
V
John
55
G
R
X
Alice
66
A
N
Z

Note the following:

  • Cells on rows containing user=Mary and a DEL-PERSON label are impacted.
  • Because of ID expansion, cells on rows containing AAID=77, AAID=88 or AAID=99 (which are the AAID values on rows containing user=Mary) and a DEL-DEVICE label are impacted. This includes cells with a DEL-DEVICE label on rows where user=Mary. This causes cells in rows 4 and 5 (as well as rows 1-3) with DEL-DEVICE labels (AAID, MyEvar2 and MyEvar3) to be obfuscated.
  • The expandIDs setting does not expand to the call to include values present in MyEvar3 (X, Y and Z), which has an ID-DEVICE label, when user=Mary. ExpandIDs only expands to include Visitor IDs (AAIDs in this example, but also the ECID) on rows where user=Mary. Thus the last two rows, which contain MyEvar3 values of X and Z are not impacted.
  • MyEvar2 in the fourth and fifth rows are updated because these rows contain the same Visitor ID values (77 and 88) as those on the first and second rows. As a result, ID expansion includes them for device-level deletes.
  • The values of MyEvar2 in rows two and five match both before and after the delete. However, after the delete they no longer match the value N that occurs in the last row, because that row was not updated as part of the delete request.
  • MyEvar3 behaves very differently than it did without ID expansion, because without ID expansion, no ID-DEVICES matched. Now, AAID matches on the first five rows.
recommendation-more-help
2969e653-1f9b-4947-8b90-367efb66d529