Ad Ops: 5 Tips For Running A Successful Ad Campaign


A successful ad campaign depends on many factors. Some of these factors—such as the time of year, the economic climate or current events—are outside of our control. While these factors could make or break the return on investment (ROI) for an ad campaign, programmatic advertising allows us to focus on the factors we can control, which helps ensure that our campaigns run as effectively as possible.

Once your campaigns have been set up, keep these five tips in mind while the campaigns are spending to maximize results and increase team transparency along the way.

After A Campaign Launch

1. Give it time.

After launch, it can be tempting to make changes to the campaign within the first 24 hours. For example, maybe you have a CPA (cost per acquisition) goal of $20 and after one day the campaign is showing a $2,000 CPA. It can be tempting to immediately try optimizing the campaign by blacklisting all sites that have delivered impressions without actions, or removing creatives and exchanges which aren’t hitting that $20 goal. However, manually “optimizing” so soon will restrict the campaign’s ability to spend in the future, and stop sites and creatives from potentially delivering conversions down the road.

Some DSPs, including the DataXu platform, build automated learnings through artificial intelligence, which takes time. Give the campaign a chance to gather enough data to make informed decisions before implementing changes.

2. Test only a few methods at a time.

Remember the scientific method? It states that, when running experiments to test a hypothesis, the best way to discover which variables affect change is to test one variable at a time. Though you won’t find test tubes or Bunsen burners in most Ad Ops departments, the principle holds true for programmatic advertising.

Let’s say you have a campaign that needs to hit a certain CTR (click through rate) goal and a certain viewability score while remaining under a specific CPM (cost per thousand impressions).  It might be tempting to take 10 days of performance data and implement a number of optimizations on day 11, such as blacklisting, whitelisting, moving budget into retargeting, removing creatives that are not driving actions, raising bids and building new test flights. However, when a dramatic performance change shows on day 12, how do you know which campaign modification affected your results?

In testing fewer variables at a time, we’re able to identify which adjustments drive success and which don’t, all of which are considered in the following rounds of optimizations.

3. Talk about what you’re doing.

In a detail-oriented industry like programmatic advertising, it can be easy to get bogged down in the weeds of your daily tasks. However, at DataXu, it’s common to see people out of their chairs, collaborating with others. We share experiences and ideas not just within Ad Ops, but across different departments. For example, it’s not unusual for people from Ad Ops, Sales, Optimization, and Engineering to be engaged in conversation around a table, solving a specific problem or developing ideas for new technologies.

Ultimately, everyone on the team has a perspective on how to create success. For us, the “team” goes beyond the DataXu offices and extends to the advertiser, the agency, and the external ad servers and partners who help us measure and qualify our data. The more we share our insights with each other, the better we’re able to move in the right direction.

4. What works for one campaign won’t always work for another.

Airlines, automobile companies, streaming services and insurance companies represent four different industries we work with at DataXu, each with its own business model and unique metrics for success. Even for a single advertiser, we recognize that there can be multiple methods to reach that advertiser’s marketing goals. The safest way to ensure optimal performance for a campaign is to take the time to determine what will work best for each specific campaign, no matter the industry or advertiser.

For example, an individual domain might perform poorly on desktop, but perform very well on mobile or video. One advertiser may see a specific call to action or creative size result in a higher ROI in one initiative than in another. The recipe for success may vary depending on the goals for a specific campaign. Optimizing each campaign individually keeps it from going down the wrong track.

5. Take time to evaluate the data before you try something new.

Make sure that the data you’re reviewing is recent and specific to the advertiser, campaign and channel you’re evaluating. Using timely, relevant facts to form optimization plans keeps you in control of campaign performance. For example, just because one creative size performed poorly on campaigns in the past, review its current performance before potentially removing it. And once all of the data has been weighed and you make a change, evaluate the data again to see whether the hypothesis was correct.

Understand and Adapt

Continuous improvements to programmatic technology give us the power to understand who is responding to our campaigns, when and where they’re viewing it, and the devices they’re using to engage with a brand. The insights we gain on a regular basis mean the paths to success are evolving constantly. In short – the more understanding we have of how the entire campaign is running, the better we’re able to adapt to those factors which are out of our control.

While this list is by no means comprehensive, we hope these tips serve as a helpful guide. To learn more about setting up your company for programmatic success, download our 2017 Marketer’s Guidebook on How To Build A Programmatic Organization here.

Do you have a technique that’s worked well for you in running successful campaigns? Let us know on LinkedIn, Twitter, or Facebook.


RELATED BLOG POST: 5 Ad Ops Tips for Building a Successful Campaign