Crosswalk and taxonomy quick start

14 min

overview this article defines how to connect segment identifiers to make the appropriate household associations we refer to this as the crosswalk to enable data partner segments in audience studio marketplace, two things are required establishing a crosswalk / id sync between cadent and the data provider (you) setting up a taxonomy of segments for the provider, which will appear in audience studio and give advertisers the ability to create audiences for campaigns using your segments workflow the following defines the process for enabling data partner segments set up two destination points one at cadent and one with you as the partner the cadent endpoint receives the file of identifiers from you, matches them to the cadent identity graph, applies a synthetic id, and returns a file of data partner ids with your associated synthetic id you and cadent use the synthetic id's to share segments the file exchange uses sftp or an s3 buckets to store and retrieve data send cadent a file containing the ids this first file, referred to as your data file or crosswalk file, should consist of the agreed on primary partner ids, along with all associated segment ids for example you may have one primary partner data id in multiple segment ids optionally, you can include a second or even a third identifier tied to each of your segment ids, by including additional columns for the added ids if you don't have your own proprietary ids, you can send another identifier in its place if necessary, you can send multiple ids (for example, maid and a zip 11) these can contain any of the following zip 11 (preferred) postal address mobile advertising identifiers (maid) ip address using ipv4 or ipv6 hashed email address (hem) note if you, as the data partner, have your own proprietary ids, cadent can use that id as the main identifier for data sharing cadent will need your proprietary id tied to one or many of the identifiers listed here to make the appropriate household associations we refer to this as the crosswalk in certain instances, cadent may send you its data points, for you to assign your own ids to them send cadent a taxonomy file containing segments the second file , referred to as the taxonomy file, should be your most recent taxonomy or segment listing if you plan to use custom taxonomies, speak with your cadent representative because this article does not cover these after the first time a crosswalk is established, you may need to send a third file, that includes a mapping of your old phased out segment ids to the new segment ids after cadent maps your sunset segments to your new segments, there may be remaining segments from the old taxonomy for which you don’t have a new equivalent cadent will inform any partners using such segments that these are about to go away, and give them a period of approximately two weeks to adjust their campaigns accordingly if you have any segments with a household count of 0, cadent will notify you for a chance to review segments with a count of 0 are excluded from the audience studio data marketplace file formatting crosswalk file formatting this section defines the formatting and contains the minimum requirements for submitting your initial or updated crosswalk file cadent sends you all options in a separate csv template this file is a mapping of your primary partner id (or another type of id) and a list of segments where this id falls the following table provides an example \partner id or other id\ \secondary id zip 11, maid, ip, or hem\ \segment id\ 12345 27604 2017 01 segment 1,segment 2,segme nt 3 67890 27604 2017 03 segment 1 the preferred file format is csv following are the preferred delimiters pipes following the partner id columns commas for the segment column you can include a partner id column for proprietary id’s and a matching data point column named for that data point (for example, zip 11, ip, or maid) you can enter multiple data points in individual columns optional format use this format only if the preferred one is not available to you enter one file per segment the segment id must be in the filename; cadent will extract it the file content is a single id column \partner id or other id\ 12345 67890 taxonomy file formatting ensure that the file contains a mapping of the segment ids that appear in the crosswalk file the filename should clearly identify the file as a taxonomy and not a data file we recommend a format that consists of the name of the audience, the taxonomy name, and the date the file was created, as shown in the following example audience name taxonomy name date csv the minimum requirements for a taxonomy file consist of three columns, provided there is no nested hierarchy in these segments \segment id\ \segment name\ \cpm\ segment 1 frequent paper towel users $1 75 segment 2 infrequent paper towel users $1 25 complete format options will be sent to you in a separate csv template in cases of a nested hierarchy, following are the minimum four columns \segment id\ \parent id\ \segment name\ \cpm\ category name 1 root goods subcategory name 1 category name 1 paper towels subcategory 1 seg ment 1 subcategory name 1 frequent paper towel users $1 75 subcategory 2 seg ment 2 subcategory name 1 infrequent paper towel users $1 25 taxonomy file formatting best practices the following practices are required for the taxonomy files the segment ids are the same as the comma delimited ids from your crosswalk file only alphanumeric or underscore characters are allowed within the file do not use the following characters !, @, #, $, %, \& , ^ the length of each field should not exceed 1024 characters if you have special characters in your segment name, enclose the name in double quotes send file updates send a follow up file if you are sending additional segments (not for a full refresh) for example, if you submitted a file for three additional seasonal segments after we already did a full quarterly refresh now, we need to know which partner ids fit into these new segments, so we need to update the data as well the format to create the additional crosswalk table is defined in the crosswalk file formatting section \partner id or other identifier\ \segment id\ 12345 segment 1,segment 2,segment 3 67890 segment 1 taxonomy mapping file format in cases where your old taxonomy contains segments that do not appear in the new file, send cadent a mapping file of old to new segments to keep the campaigns using old segments running without interruption you can deposit this mapping to the same endpoint as the crosswalk and taxonomy files identify it as a mapping file in the filename the following table shows an example \expired segment id\ \new segment id\ 12345 67890 abcde fghij cadent will send you back a list of segments that are outdated and have no new segments mapped, so you can review and confirm the data cadent then informs any partners using the segments and gives them a grace period of two weeks to make changes data submission after pushing the files to the endpoint, send an email to your cadent representative with the following recommended information subject \<data partner name> \<dataset name> to enter multiple files, enter the description for each unique file location on a separate line the email body should contain the following filenames \<file locations on s3/sftp> \<file locations on s3/sftp> taxonomy file for reference, the full taxonomy file is available here for reference in excel format