Broadcast Filters - Advanced Options

This article applies to Broadcast touchpoints, whether using our full funnel LeadsRx Attribution or our LeadsRx Broadcast Attribution.

By default, any visits to your website being tracked within 10 minutes of a broadcast spot running will be attributed to Broadcast. Your broadcast touchpoint will receive credit for any conversion that visitor achieves in the future. There are other filters you can place on this data to narrow the scope of what credit goes to broadcast.

Geo-fence:

If you are running advertising in local markets, we highly recommend geo-fencing your broadcast data. Geo-fencing will only analyze visitors who are located in the market that is being advertised. For example, if you are running a spot in New York, you would not want someone who visits the site within the attribution window from LA to be attributed to that spot, as they could not have been influenced by it.

Geo-fencing is available at 3 levels: DMA (default), city and state, or zip code depending on your marketing strategy. The setting for how to track geography should be decided on when you first open an account and is set during the onboarding process. The default for new accounts is to use DMA as the methodology.

For steps to implement Geo-fencing, please check out these articles depending on which LeadsRx product you use.

• LeadsRx Attribution (using the LeadsRx Universal Pixel) - How to Geofence Broadcast Touchpoints in the Full LeadsRx Attribution product
• LeadsRx Broadcast Attribution (relying on data from Google Analytics or another analytics product) - Geo-fencing Broadcast Web Sessions

Channel Filters:

When someone is exposed to a broadcast spot, the two ways they will most likely visit your site is through either a direct visit or via a search engine. Channel fencing will allow you to exclude any visits that came from channels that are unlikely to have been influenced by a broadcast spot.

We recommend given broadcast credit for any visits from either direct traffic or organic search. You probably would not, however, want to take credit for a visit via a paid social channel, an email campaign, etc. These visits were unlikely to have been influenced by broadcast.

The decision whether to broadcast credit for visits that come in via paid search is more of a judgement call. This decision relies on your business model.

If you are running a paid search campaign alongside the broadcast campaign, then you may not want to cannibalize visits from paid search to apply them to broadcast. However, if someone does do a search after seeing or hearing an ad, the first search result they see is likely to be the paid search ad. In this situation, it is valid to give broadcast credit for these visits.

One compromise option is to designate branded vs. non branded search in the touchpoint being ingested, and then only take credit for the branded paid searches, not the non branded. To learn more about distinguishing between branded and non-branded paid search, read Creating an LRX Tracking Parameter for a Custom Touchpoint.

You can have broadcast data overwrite at the touchpoint or grouping level. We recommend overwriting at the grouping level to save time. That way, when you keep your touchpoints in the appropriate groupings, that is also organizing your broadcast targeting.

Channel Fencing is set within the Advanced Options area of your broadcast touchpoint or campaign.

Attribution Window and Decay Curve:

While by default the attribution window is set to 10 minutes, this can be adjusted. LeadsRx has a tool called the decay curve that will assist in determining if this is appropriate.

Once a significant amount of spots have been loaded into LeadsRx (this varies depending on how much website activity there is), you will be able to analyze when people are visiting the website on aggregate. Below is a good representation of a decay curve.

We almost always see this type of visitor behavior where there is a spike early on after the spot ran and then the visits declined back to the average level (as measured by the 5 minutes before the ad aired) This is what is represented by the white horizontal line. This most commonly happens just before 10 minutes after the spot ran. This example curve shows why we use a 10 minute window by default. However, if the spike in your decay curve happens at a different time, then you can adjust your attribution window to make it longer or shorter. This is done in the advanced settings section of the broadcast edit menu.

Prediction Period (Lift Report Only):

Unless you indicate a predictive model to be used, LeadsRx will consider 100% of website visitors who meet all of your filter criteria and fall within your attribution window, as having been potentially exposed to your broadcast advertising. No baseline is used for attribution by default.

If you choose, LeadsRx can use a predictive model to determine expected web traffic. Only traffic above the model prediction will be included in attribution results. This model is based on a number of weeks prior to the start of broadcast advertising and looks for directional change over that time. This becomes your baseline.

NOTE: Setting a longer prediction period may help overcome anomalies due to seasonality or special events. If for some reason, traffic to the website was higher before broadcast began running, applying a prediction period may actually result in no attribution being given to broadcast. Take this into consideration when deciding whether to use this feature.

The predictive model makes the most sense if there has not been any broadcast advertising running during the predictive period. You can use a predictive model of 2, 4, or 6 weeks before your touchpoints spots begin.

Responder Ratio (Lift Report Only):

The Responder Ratio represents the perceived strength of your broadcast advertising. By default, broadcast touchpoints start with a responder ratio of 100.  This means that 100% of the visitor traffic during the Response Window (that meets all your other filter criteria) should be attributed to having seen or heard the broadcast advertisement. Setting this value to a lower number will reduce your calculated broadcast lift accordingly. The responder ratio does not impact the impressions or conversions recorded for your spots, so your segmentation reports will not change. This is because broadcast attribution uses a probabilistic model while interactions and conversions use a deterministic model. Read Probabilistic Versus Deterministic Attribution to learn more. 

Many LeadsRx clients will leave this value at 100 and use the other advanced settings to more narrowly define which visits are attributed to broadcast.