The Ins and Outs of Digital Ad Trafficking and Ad Ops
For this article, we hired a top-secret Ad Trafficker pro to give us the scoop. He or she has asked to not be named to protect the innocent.
What is digital ad trafficking
A digital ad trafficking campaign builds out a media plan into a live campaign that serves and tracks ads. Create, prepare, and execute digital advertisements (banner ads, programmatic, rich media, etc.) to launch a digital advertising campaign. A digital ad trafficker may set up tracking parameters, schedule an ad to launch at a specific date and time, and ensure that it runs correctly in the right places, regarding the creation of a new creative container for an ad.
Digital ad trafficking consists of three primary components: creative assets provided by the creative team, trafficking instructions provided by the media team, and ad server configuration and tag generation provided by the ad operations team.
Ultimately, ads are served across the digital ecosystem and tracked in media reporting systems by the creative, media, and ad operations teams. Meta data is generated and appended to each ad during the configuration stage. For reporting and optimization purposes, this allows media teams to analyze ads at a granular level. Generating and appending this essential meta data remains almost entirely a manual process. As a result, advertisers rely on labor-intensive, insufficient, or both ad-hoc forms of media metadata governance.
Did you know? Claravine is solving ad trafficking headaches for the largest brands and agencies in the world.
Key stakeholders
As previously mentioned, there are three primary stakeholders in the trafficking process: the creative team, the media team and the ad operations team.
Creative Team
The creative team is responsible for providing finished, functional, to-spec creative assets to the media team. The process begins when assets are passed to the media team or loaded into the adserver by the creative team. The media team then passes those assets along to ad operations or passes along unique identifiers to locate the creatives if they have already been uploaded to the adserver. If assets are out of focus, contain misspellings, do not have borders, do not meet specs to load properly into the adserver etc. then the creative team must revise accordingly. Once assets have been passed by the creative team to the media team or into the adserver, the creative team plays an advisory role should any issues arise. This is until the trafficking process has been successfully executed.
It is worth noting that there are two primary methods of trafficking creatives. In the past, the creative team would deliver the creative asset to the media team. The media team would then provide it to the ad operations team along with completed trafficking instructions. Today, as more advertisers demand rich media functionality across their ad content, more and more campaign managers are using solutions such as Google Studio to create and send assets directly to client ad servers for trafficking.
Media Team
The media team is responsible for ad campaign and placement—individual ad—creation in Prisma (campaign management platform), along with providing the trafficking instructions document to the ad operations team. This is where the bulk of the work and thus, room for error exists in the trafficking process. As detailed below, creating campaigns/placements along with the correlated trafficking instructions document is a tedious, laborious process that is prone to human error. This is also the most critical step in the process; any errors here lead to cascading issues downstream in analytics workflows.
Ad Operations
The ad operations team is responsible for implementing the trafficking instructions provided by the media team in the ad server. These instructions are used to generate tags that are distributed to a publisher or appended to a buying platform. The ad operations team must operate under the assumption that the trafficking instructions they are provided are complete and accurate; this is the media team’s responsibility.
The Ad Trafficking Process
Now that the media campaign and the creative strategy have been planned and funded, the activation is ready for execution. Steps include:
-
The creative team generates the assets and either sends the assets to the media team or loads the assets into the adserver (reference above for two main workstreams)
-
The media team creates the campaign in Prisma according to a predetermined taxonomy
-
As part of the campaign, the media team creates placements in Prisma in accordance with the media plan while adhering to a predetermined taxonomy. There are typically hundreds to thousands of placements per campaign, and potentially hundreds of campaigns per advertiser per year. Each individual placement must be created and configured manually. This is tedious, error-prone work that has proven extremely difficult to properly govern for many advertisers.
-
Media team pushes campaign with placements from Prisma to adserver
-
Media team exports newly added campaigns and placements, along with correlated IDs, from the adserver to the trafficking instructions doc for ad operations. Additionally, the media team will define which creative assets are assigned to each placement in the trafficking instructions. The ad operations team will use these trafficking instructions to configure each placement in the adserver (i.e. Google Ad Manager). Fields that require configuration include, but are not limited to:
-
Flight dates
-
Creative dimension
-
Creative name
-
Creative rotation
-
[crs’ vary drastically so pull some more fields out of whatever crs you display in this blog article and list here if desired]
-
-
Media team provides said instructions (along with creative assets if necessary) to the ad operations team for implementation in the adserver
-
The ad operations team implements the trafficking instructions per the media team’s direction, generates ad tags, and distributes those tags to publishers and/or media teams to implement in buy platforms (DV360, TTD, Amazon DSP, VerizonMedia DSP, etc.) for activation
How do we help with ad trafficking
Watch this interactive demo to see how we help you standardize and connect campaign, placement and creative data
Pitfalls in ad trafficking
The primary pitfalls with digital ad trafficking come when dealing with large advertisers that demand the granularity required to perform advanced media analytics. Extensive, granular tagging for large, national advertisers necessitates media teams manually generating potentially tens of thousands of distinct placements a year.
What often happens is that incorrect metadata is created for placements as depicted above. Because there is no proper system of data governance in place, the erroneous placements are trafficked, tags are created, and the media is run against these tags. From here there are two outcomes:
-
The media analytics team does not catch the mistake among the many thousands of distinct placements in the dataset they are analyzing and unwittingly produces invalid results. This is where the issues cascade; a DSP optimization algorithm could be informing media bidding strategy using flawed data, MTA and MM models may be informing short- and long-term investment planning using flawed data, conversion pathing analysis could be grossly misleading, on-target percentage scores would be off, optimal frequency calculations under or over reported. In short, any data and analytics workstream will be fundamentally unreliable.
-
The media analytics team realizes a mistake was made during placement generation, at which point it is impossible to perform the desired data-driven task.
Tips to manage and improve digital ad trafficking steps
Orders and line item data must match the IO exactly, including naming conventions, values, etc.
How do we manage/improve this?
Implement a cloud-based system that matches your orders and line item data that directly connects to the ad server. The key problem here is that order data and line item data differ across all buying platforms. There are no uniform inputs. You can build platform-specific connectors with built-in data governance features, but the ecosystem is very fragmented.
Task: When running multiple indirect partners (ad networks, SSPs, exchanges) check network settings, priorities, passback scheme, frequency capping, & CPMs.
How do we manage/improve this?
Media ‘traders’ are constantly monitoring these variables and more for live biddable media campaigns. Buying platforms are constantly improving their ML to automate and/or improve campaign monitoring and optimization.
Task: Put all updated creative/tags in a test environment that mimics your live one, for thorough QA.
How do we manage/improve this?
Instead of mimicking your live environment, install a tag monitoring system on your production site. Uplevel by having a system that does automated, routine checks with email alerts when something fails. Many AdOps teams do this in Google Marketing Platform (formerly DCM).
Task: After launch, check the ad server within the first 90 minutes to confirm impressions are served & clicks are tracking.
How do we manage/improve this?
Media teams we’ve worked with do this in GMP(DCM) and the agency analytics teams we’ve worked with do this in Google’s Data Transfer files product (updated daily)
Consider also using a calendar or your project management tool to set up a reminder that includes notifications to the service or services (e.g. SMS or Slack) that you are most likely going to see. Slack has lots of helpful integrations with other tools like Google Calendar, Asana, Trello, etc.
Task: Check 3rd party system, at least 24 hours after launch, and compare deliverable data with 1st party ad server.
How do we manage/improve this?
This depends on your data infrastructure; if for instance your 3rd party system is dropping data into your data environment (via FTP through an autoML tool into a db like many set-ups) then this type of QA is feasible and can be semi-automated (media team provides analytics team with a running mapping document of campaign identifiers and launch status, analytics team hosts that data in a table, query is automated to check for delivery data for campaigns with launch status live and can flag issues via automated email alerts or a live dashboard) but that’s a pretty heavy lift, would only be valuable for advertisers with complex media strategies.
Task: Take screenshots and send them to the respective advertiser or agency immediately following launch.
How do we manage/improve this?
Techsmith’s Snagit is a helpful tool for this. While screenshots may be a handy way to report back, we actually recommend recording a quick screen share video. This is because you can highlight other elements of the page including scrolling the full width, interactions, tags firing properly, etc.
Also, create a shareable directory of all screenshare videos and/or screenshots related to the team you’re sharing it with, properly naming files with respective campaign attributes and dates.
Task: Create automated reports for 1st and 3rd parties each week.
How do we manage/improve this?
Besides weekly reports, real-time notifications related to performance are recommended. Example 1, setting up continuous monitoring and getting notifications of tag/page validations.
However, you probably also require an army of engineers and data scientists … the type of event and user level reporting and media optimization capabilities that sophisticated clients are increasingly demanding are extremely hairy big data problems considering the fragmented ecosystem, the business model driven by the walled gardens, the developments in eprivacy regulations, and the sheer volume and variety of data that adtech systems are generating. Be proactive in pointing out campaign performance “highlights” and “lowlights” to the advertiser/agency and providing suggested optimizations.
Thanks for reading! If you want to learn more about how Claravine helps in the digital ad trafficking process, get in touch. Not quite ready to chat? Check out our guide on automating media ops for better analytics.