By Frank Totera – Atlas Technical Support Team |
Welcome to the latest installment in our Best Practices blog series. This series is designed to help our customers get the most out of Atlas — from building media plans and trafficking campaigns to measuring and understanding results.
We’re taking a deeper dive today, outlining steps you can take to manage your report data. This discussion will be most applicable for clients who are already familiar with building campaign reports in Atlas. It’s important to understand how your data is displayed in the reports you pull, as well as the ways it can change.
Here are a few important things to keep in mind to ensure you’re building your reports the way you want them:
• Your date range will determine your data.
This may seem obvious on the surface, but with so many factors involved, it can be easy to forget. For example, in instances where a conversion occurs within the requested time frame but the impression and click occur prior, only the conversion will be depicted — every data column aligns to your date range.
• Results may show up on more than one line.
Atlas will separate your data into individual line items where possible, based on your current level of aggregation. For example, impressions, clicks and conversions will be broken out on different lines in instances where each metric is given a unique level of aggregation. The unique level of aggregation is determined by the dimensions included in your report definition. An Atlas client who wants to include the click-through URL dimensional column for ads with multiple click-through URLs would see each one reported individually. Similarly, a client pulling multiple actions that have received conversions against the same placement would see these conversions on separate lines.
• Conversions may be reallocated.
Have you ever noticed that your conversions change slightly between report outputs? This is because Atlas attribution is dynamic. In other words, conversion processing is completely run at report time and there is no static table storing conversion data. Because of this, Atlas reports may reallocate conversion credit if certain changes are made. Here are some factors that can affect the conversion output in reporting.
• Changes to conversion windows.
Changes to the view/click conversion windows for an action will alter how conversions are processed, and the new report will display conversion counts for the action based on the new conversion windows . For example, an action with a one-day view window would not report a conversion from a person who waited two days after seeing the advertiser’s ad to trigger it. But if the view-conversion window was later updated to capture three days, the conversion would show in your report.
• Changes to the advertiser the action is shared to.
Sharing actions to additional advertisers may cause conversions to be reallocated to impressions and clicks from these advertisers if they supersede impressions and clicks from other advertisers the action is shared to. For example, if a person clicks on ads from both advertiser A and advertiser B prior to triggering an action shared only to advertiser A, Atlas would allocate the conversion to advertiser A in a last-touch attribution model. However, if the action was later shared to advertiser B, it’s possible that the conversion would now be reallocated to advertiser B instead if the click from advertiser B occurred after the click from advertiser A.
• Changes to advertisers in the action attribution filter.
Atlas gives clients the option to calculate their conversions against a subset of advertisers that an action is shared to. This checkbox can be found beneath the actions filter in the report form. If the action attribution filter is unchecked, Atlas calculates conversions against all advertisers the action is shared to by default. If attribution filtering is selected, only impressions and clicks from the selected advertisers are included in conversion processing, while the impressions and clicks from the advertisers not included in the filter will be ignored. For example, suppose a person has clicked on ads from advertiser A and advertiser B prior to triggering an action shared to both advertisers. Atlas would normally attribute this conversion to advertiser B in a last-touch attribution model. However, assuming that the subset filter was enabled for this action — and advertiser B was excluded from it — Atlas would attribute the conversion to advertiser A instead.
And that wraps us up! Now that you’re well equipped to tackle campaign reporting, you can leverage the full range of benefits Atlas has to offer. As always, if you’re a current Atlas customer with questions about reporting or any feature, you can visit the Help Center or contact your Platform Solution Consultant for support.