Every ad that runs on a site serves different goals. The advertiser’s goal is to generate sales; the publisher’s goal is to generate revenue. But most often overlooked is the goal of the user.

The assumption, unfortunately, is that most every time you run an ad, you’re going to bother the user. There are rare exceptions to this - occasions in which you’ve delivered the right ad to the right person at the right time, the user engages and ultimately clicks through. But that happens only six out of every 10000 times the ad is served – and that’s a generous estimate.

The bulk of the responsibility for addressing this dilemma falls on the publisher, who needs to find ways to optimize not only revenue, but also the user experience. How can this be accomplished?  Rather than piling on more and more ads, publishers should consider delivering a single high-impact format – one that potentially increases the length of time the user spends with the ad and results in a more positive interaction overall.

Traditional high-impact formats include pre-roll, in which the user is forced to watch a 15 or 30-second ad before he or she can view the content they’re really after, and interstitials, which temporarily take over the page your user is trying to get to. With these formats, you’re likely to get his or her attention – but will it be the kind of attention you want?

Compounding the problem is that unless a publisher is selling all of its inventory directly, it has little control over what ad will be shown at a given time, which means it could potentially be an ad with offensive language or unpleasant visuals. But the publisher does have the ability to control how the ads behave, such as where it appears, for how long, whether or not the user can easily dismiss it.

Sadly, there’s no easy fix to managing all the variables of how an ad behaves, and determining which combination of variables works best will vary from site to site. Publishers clearly need better ways to manage and optimize all the options, rather than more complexity to manage.  Therefore, it’s incumbent on publishers to test, test, and then test some more, then analyze their metrics to determine what results in engagement or abandonment, to determine what methods are most likely to work best in the future for various scenarios.

Here are the questions every publisher needs to answer in an effort to better manage the complexity:

- Audio: Sound is one of the most critical sources of arguments today between advertisers and publishers. In short, advertisers always want the default setting to be audio-on, while publishers worry about unwanted sound intruding on the user’s experience and, at worst, driving their audience away.

- Viewability: Does it make sense to show a premium ad in a very dominant location for a few seconds, then have it minimize to a less intrusive place if the user does not engage immediately? What ad placements work best, and why? Are adhesive or sticky ads – ones that remain at the bottom of a mobile or desktop screen while the user scrolls – more effective than static ones?

- Skippable Ads: How much freedom of choice can you provide users while still serving your own interests and that of the advertiser? Does allowing users to dismiss, or skip the ads they don’t want to see make them return more often and stay longer – and do they use that time to engage with ads that fit their interests?

- Revenue: Do fewer, higher-impact ads work better? If so, can you use your metrics to charge premium rates for these ads? Can you do without, or severely curtail, the use of lower-impact ads that tend to clutter your site? How does dominant placement of a premium ad followed by a less intrusive placement impact performance?

This kind of intensive, internal analysis needs to be done on a per-session, not per-unit level. The task may seem daunting, but even the most successful publishers – and advertisers and consumers, for that matter – need to continually review, refine, and reinvent. A publisher who is able to indirectly measure the advertiser’s benefit using the effective CPM they are willing to pay over time, now needs to define the user’s benefit, through the length of the session, and through pages visited. Only looking at these two parameters compounded will reveal the best strategy for the publisher. If done well, publishers will be far better equipped to manage the complexity of revenue generation and user experience.