Violation Prevention Logic
The Challenge
As part of our Ad Operations and technical services to our clients, we make sure contractual agreements are not violated. This involves auditing, monitoring and improving ad logic when we detect that we are at risk of violation. In this specific case, we came across a challenge in which commercial elements were stacking at the bottom of the page, clashing with a specific guaranteed views commercial element.
The article pages on all the portfolio’s sites were configured so that the last commercial element on the page was a widget for a specific client. This widget had to get guaranteed views as part of the contractual agreement. This element was governed by a set of rules that could impose a fine if violated and the revenue generated from this was significant.
ADOPSTAR'S RESPONSE
The main rules of the agreement were that nothing of commercial, editorial or internal use should be directly above the guaranteed view widget. That included anything that could cause the user to leave the page before scrolling down far enough to get the guaranteed widget in view (such as redirects, links to other sites within the portfolio, affiliate links or even links to government websites related to the article content). Moreover, the rules of this agreement were very strict, outlining that even a single violation on a single article would count as a contract breach and impose the risk of a fine.
Taking into account that the portfolio consisted of thousands of news articles, with hundreds of new articles created every day on 70 websites, the need for an automated solution was urgent. If the checks for violations and optimisation on high-risk content were left to be resolved manually, the resource needed would be huge and there would still be a risk of human error that could jeopardise the whole effort. In a typical month there were about 600 high-risk articles across the portfolio.
Adopstar’s solution was to create a piece of code logic that would apply to each page during the initial load. In the case that a “violating” element was detected within a certain distance of the guaranteed views widget, then this element would be automatically removed. That distance was set to a minimum of 200 characters, which — on a typical article on those sites — would expand to a minimum of three lines of content, which was the main rule of the contract.
the result
As a result, the risk of violations was eliminated, as the applied solution ensured that there would be at least 200 characters of “clean” content above the widget (no ads, links, other widgets etc). The second part of the project was to create a detection API that would run automatically on all the articles on all sites and detect any extreme cases that may still pose a risk (for example an article with total content of less than 200 characters). These were rare and hard to find manually, but with the use of the API, all such articles were automatically added to a Google Sheet and could then be optimised individually. After the solution was applied, the number of high-risk articles across the portfolio dropped to 10-12 a month and the resources required to amend those were minimal.
Automating the API to run on a server in the early hours of every day meant that the list was ready by the time the client’s teams were online in the morning and any risky articles could be amended quickly and easily.
For more information or to find out what we can do for you, email:[email protected]
