What discrepancies should I expect?

As you implement Advanced Tracking impression and click tags across your active partners and channels, there will always be some activity that is either missed or not populating within your dashboard.

In order to give you the tools to understand just how much of the activity you are missing, we help calculate a metric we call Tag Health.

Tag Health

We calculate Tag Health by taking the data ingested via an API connection with your active channels and compare it to the number of tracked activities we capture with our click and impression tags.

The comparison of these two metrics: AdRoll Tracked vs. Channel Reported, gives us a % of coverage for your marketing activity. For example, if Facebook reported 100 clicks over the last 7 days, and AdRoll tracked 88 clicks, the Tag Health would be 88% (88 AdRoll Tracked / 100 Channel Reported). 

What is good Tag Health?

Understanding that there will always be discrepancies should help you begin to understand what good Tag Health is. Within our product, if >70% of activity is being accurately tracked, we consider that good health. What that means: You are missing ~30% of activity. So if you’d like to improve that coverage, and aim for a higher Tag Health, look at the individual campaigns or channels that you can improve.

Keep in mind: There is an expected discrepancy between 5-15% no matter how perfect your implementation is. So if you have >85% Tag Health, you are doing GREAT. 

How do I look at my Tag Health?

Navigate to your AdRoll Dashboard and find the page labeled “Tags” under Cross-Channel Attribution. Within that page, you will see your overall Tag Health score, across all connected channels, as well as a breakdown of Tag Health by Channel and Campaign.

You can always export this data to analyze on your own, simply click Export and select the level of granularity you’d like to see (Channel, Campaign, Ad Group, Ad). 

What are some reasons for data discrepancies?

Tag Health: Likely some activity is missing tags. But even if no tags are missing, it’s expected to have a modest percentage of ads, emails, texts, or assets served, that encounter issues firing our tags properly, thus missing out on the tracking of some touchpoints. 

Daily Reported Impressions/Clicks: For channel-reported activity, we ingest impression, click, email, and touchpoint data directly from the channel itself. If any discrepancies are found, they are likely due to: 

  1. Time Windows/Time Zones - our data is pulled daily and will represent UTC while some platforms will adjust the daily reporting to reflect a localized setting.
  2. If a sub-selection of accounts or campaigns has been applied using our admin-only account selector, you might see differences in the AdRoll ingested vs. channel-reported deliveries. 

Daily Reported Spend: For channel-reported activity, we ingest spend data directly from the channel itself. If any discrepancies are found, they are likely due to: 

  1. Time Windows/Time Zones - our data is pulled daily and will represent UTC while some platforms will adjust the daily reporting to reflect a localized setting.
  2. If a sub-selection of accounts or campaigns has been applied using our admin-only account selector, you might see differences in the AdRoll ingested vs. channel-reported daily spend. 

Total Attributed Conversions by Channel: Remember, one of the reasons you adopted this tool was to reduce duplicated conversion reporting that occurs when a marketing mix is not able to apply a cross-channel attribution model. This means that the conversions each channel reports as ‘attributed’ will likely be (much) more than the credited conversions within your Advanced Tracking dashboard. This is expected! Your cross-channel model looks at all tracked touch points associated with the conversion and then divides credit accordingly.

Think about a customer path like this:

Google Search >> Site Visit >> AdRoll Ad >> Facebook Click >> Attentive Text >> Purchase

Each of those channels/partners will report that they drove 1 conversion. Add them all together: 1 Google + 1 AdRoll + 1 Facebook + 1 Attentive = 5 attributed conversions. But we know those 5 conversions are really just 1. So instead of giving everyone credit for the same conversion, we divide the credit according to your model: 

  • Linear: Google .25, AdRoll .25, Facebook .25, Attentive .25 = 1 conversion
  • First Touch: Google 1, AdRoll 0, Facebook 0, Attentive 0 = 1 conversion
  • Last  Touch: Google 0, AdRoll 0, Facebook 0, Attentive 1 = 1 conversion
  • Time Decay: Google .1, AdRoll .2, Facebook .3, Attentive .4 = 1 conversion
  • Positional: Google: .4, AdRoll .1, Facebook .1, Attentive .4 = 1 conversion

It is expected that each individual channel will take more credit for your conversions than will be reflected within Advanced Tracking. One of the biggest drivers of conversion discrepancies, beyond what is expected with a cross-channel attribution model, is the ability (or inability) to track impressions.

Check out our list of channels to see which ones allow us to track conversions, and which ones don’t. 

Your Total Conversions vs. Advanced Tracking Conversions: The conversions that AdRoll Advanced Tracking reports within your dedicated dashboard represent all known marketing influenced and attributed conversions. In order for the marketing influence to be known, we need to be able to track the touchpoint. 

For example, if you are running an offline out-of-home campaign, you are likely influencing some online conversions that are not represented within Advanced Tracking. This is because we don’t know that conversion, or set of conversions, was influenced by your marketing efforts. 

You also likely have conversions occurring without the influence of an online marketing touch. You’ve built your brand well, some conversions will happen organically. Those direct, or no known marketing influence, conversions will not be represented in your Attributed or Assisted Conversion metrics. 

Lastly, our reported conversions are limited to what the AdRoll pixel is able to track. For untrackable users, like those with ad blockers or those that opt-out of third party cookies, can still convert but we will not track that conversion within our system. So keep in mind that when you look at Advanced Tracking Attributed Conversions, you aren’t looking at your total website conversions, but the portion of conversions that are influenced by tracked marketing touches.

Was this article helpful?