Best Practices: Under Delivery

The Quantcast Platform is built to drive campaign performance with minimal need to make manual optimizations. This guide will provide some best practices to optimize campaigns and achieve full-budget delivery.

Monitoring Delivery / Signs of Under Delivery

When viewing the Campaigns table in the Quantcast Platform, the Delivery column will show you under delivery alerts for active Campaigns and Ad Sets.

Under Delivery is when an ad set is unable to consistently hit the set daily spend goals. For example, if a daily budget for an ad set is $1,000 per day but for 3-5 days only $800 is being delivered -- this would be a sign of Under Delivery. This article looks at how we monitor delivery in the campaigns section of the Quantcast Platform.

First, determine if your ad set or campaign is under-delivering:

1. Go to the campaign view and click Insights on the under-delivering ad set, or campaign level if Campaign Budget Optimization is on. 

2. In the Performance tab, choose the metrics Budger Planned and Budget Delivered to confirm that we are consistently hitting your spending goal.

   

3. Check Key Events to see if there have been any changes that would affect spend, such as budget, pacing schedule, creative, dates.

Checklist For Understanding Your Ad Set and Best Practices

Now that you have determined that your ad set consistently underdelivers, follow the checklist to determine what adjustments you can make to improve delivery.

Targeting

Are there any recent adjustments that have been made to targeting? Does it align with the timeline when you started to notice under delivery? We often see changes being made to ad sets that affect delivery, especially changes to Location; choosing more granular locations will give you higher precision and, accordingly, less reach for that ad set. Broadening your Locations will have the opposite effect - greater reach and lower precision.

• Lookalike Event Groups: Make sure that the conversion event group is included.

• Retargeting Event Groups: We recommend retargeting for "All Site Events" rather than limiting our system to model from one or two specific pages.

• Location (Geo-Targeting): Geo-targeting always limits the cookies we can reach, especially if your selection is limited to specific zip codes. Expand the targeting criteria to a broader level to give our model the best opportunity to deliver.

• Demographics: Demographics are a goal constraint which is looser than a filter constraint. Goals allow us to focus on delivering to the audience as specified and— consider other valuable opportunities as they arise, helping us keep our priorities on the overall outcomes you specified for your campaigns while ensuring we deliver against your full budget. With demographics, we’ll go after opportunities that match the demographics specified on your ad set but will also consider valuable opportunities that are not a strict match for all the demographics entered. This approach favors reach over precision.

• Interest-Based Audiences/Keywords: Interests, like demographics, are implemented as goals, so they don’t impact your total addressable audience, but they do impact the choices that the system can consider when figuring out whether to bid on an opportunity. When setting up an audience based on interests, broader interests will typically have a larger audience than narrow ones, so keep this in mind when choosing your topics. nHere is a help center article highlighting ad set reach and Pre-Campaign Forecasting.

Budget and Schedule

Have there been any recent adjustments made to the overall budget, pacing, or day parting?

• Budget Changes: Whenever you increase the overall budget, the ad set spend for that day and future days will adjust accordingly. We sometimes see under delivery happen on the first day the adjustment was made while our system recalibrates how much to spend.  Best practices would be to monitor delivery on the first full day of spending after the change was made.

  • Example: If you change the overall budget on Tuesday, check delivery for Wednesday.

• Pacing: Are we pacing evenly or utilizing custom pacing? If we are utilizing custom pacing, how is pacing set? Monitor that when daily spend increases our system is meeting the daily spend. If not, see how the budget can be reallocated more evenly across other days.

• Dayparting: Dayparting prevents our system from delivering at specific hours of the day throughout the week. Do we have any dayparting implemented on the ad set? Is it set to not deliver on hours that have the opportunity to bid well? For best practices, remove any dayparting on the ad set, if possible. Our system will automatically optimize and allocate spend hourly depending on forecasted bid opportunities and performance.

Creatives

What creatives are in use? Does the ad set contain all the recommended creative sizes? Are they clicking out to the appropriate landing URL? Is the landing URL leading to the conversion event? 

• Display: Ad sets running the below display creative sizes historically perform the best in the Quantcast Platform. Ensure the correct creative size is selected at the creative level and the landing URL clicks out correctly. Check if there are any creative rejections listed. Creative rejections will show up as an "!" icon next to each rejected creative with a reason and troubleshooting steps.  Finally, make sure that the ad set is running the five recommended display sizes. Here is a link to a help center article on how to Upload and Add Your Creatives.

• Video: For best practices, include videos of 6 seconds, 15 seconds, and 30 seconds.

Advanced Settings

Advanced settings allow you to customize your ad set and ensure that your desired goals are being met. However, it also acts as an additional layer/filter that limits opportunities our model is allowed to bid on.

Best Practices:

• Viewability: Lowering viewability allows our model to have more bid opportunities. We recommend setting viewability to 55% for conversion objective campaigns.

• Frequency:  Frequency will limit how often we target and deliver an impression to the same user. Historically, conversion-objective campaigns have an increased chance for conversion if a user is targeted multiple times. We recommend not setting a frequency cap, especially if the ad set was recently launched.

• Devices: Can the ad set deliver across all devices? If not, it will heavily affect inventory and available bid opportunities. We recommend running ad sets on all devices, if possible.

• Brand Safety: Brand Safety includes Integral Ad Science (IAS) and DoubleVerify (DV) pre-bid segments and site lists. Adding in pre-bid segments or allow/block lists will limit bid opportunities and could affect delivery, especially the more strict it is. For brand safety, the Quantcast Platform has a global blocklist that prevents delivery on all high-risk sites.

• IP Suppression: IP suppression allows you to prevent delivery to certain IP addresses. However, having a large IP suppression list will suppress cookies you may not intend to.

To better understand audience insights, read the help center article here: Understanding Campaign and Ad Set Audience Insights