The digital ad fraud controversy – which encompasses everything from fake clicks to compromised viewability and increased brand risk – has been an issue for marketers for several years now. But it isn’t something marketers want to spend too much time thinking about, and for good reason. Wasted advertising dollars, unsavory marketing tactics, and a legitimate degree of helplessness to fix the issue makes for unpleasant conversations.

But every few months, with remarkable regularity, a fresh incident or research study re-introduces the topic into the digital marketing discussion. This spawns a stream of articles in the mainstream business press, putting marketers back on the hot seat. The most recent development involved ads running on YouTube next to extremist videos, which lead to coverage such as “AT&T, Verizon Pull Ads From Google Over ‘Hate’ Videos” from the Wall Street Journal, and “Google Tries to Stop Ads From Appearing Next to Hate Speech” from the New York Times.

All the while, specialists continue to monitor industry activity and hone tactics aimed at helping advertisers minimize the impact of fraud. IAS is one such company. Twice a year, they publish an assessment that tracks digital media quality and the effectiveness of best practice techniques to combat fraud.

Here’s what you need to know from the research:

• “It’s Time For Action.” Or so says Marc Pritchard, the Chief Brand Officer at P&G. The second half of 2016 brought a host of media quality issues to the forefront of marketers’ minds with fake news, extremist sites, and “Methbot” dominating the headlines. Transparency and trust will continue to be key issues in digital advertising this year.

• Proactive Management Works. IAS research showed significant performance variation between campaigns using ad fraud prevention (1.6% fraud incidence) and those that did not (13.2%).

• Timing Matters. The risk of ad fraud increases in the late hours of the night, while people are sleeping. Ad fraud begins to climb at 12:00 a.m., peaks at 4:00 a.m., then declines slowly till it reaches a steady level at 8:00 a.m. and throughout the day.

• Brand Safety Improved, But Remains a Focus. The volume of brand safety infractions decreased compared to H1 2016, from 9.5% to 8.6% of impressions on all buy types flagged on objectionable content in H2. However, with the advent of fake news and the increase in violent and extremist content, brand safety remains a critical issue for advertisers.

• Video Viewability Increased; May Invite More Fraudsters.  Considered more effective than display advertising, video ads continue to grow in demand. As such, video CPMs and subsequent expectations of quality are high. Over a sample of six billion impressions, video viewability showed significant improvement compared to H1, increasing from 40.0% overall to 58.2%, and outperformed desktop display viewability for the first time. Still, advertisers need to keep close watch on video campaigns as the higher CPMs make the medium more susceptible to fraudsters.

Tim Bourgeois (@ChiefDigOfficer) is a partner at East Coast Catalyst, a Boston-based digital consulting company specializing in strategic roadmaps, digital marketing audits, and online marketing optimization programs.