Attribution App uses a proprietary model to give you precise metrics for your return on ad spend. Below we will dive deeper into the technicalities of how Attribution App derives return on ad spend for your channels. The two primary features to understand are 'multi-touch' attribution and the 'look forward' window.
Attribution app is designed to be a multi-touch tool, this means Attribution app will consider all your marketing channels when calculating return on ad spend.
How is it calculated?
After you have properly installed the Attribution app snippet on your webpages
a cookie will be automatically be installed in the HTML of your visitor’s browser.
The cookie will allow Attribution app to see the specific visitor’s journey which includes their page views, ad interactions, utm tags, and referring links.
Attribution will use these interactions in conjunction with your desired settings to calculate return on ad spend for your ads in a specific date range.
Understanding the visitor’s journey & events.
Imagine that the black line below represents the visitor’s journey from January 1st to January 31st 2019.
Along that journey your visitor will interact with your ad channels, directly visit your website, sign in or convert. All of these interactions are considered events in Attribution App.
When you integrate your conversion tool like Hubspot or Salesforce for example Attribution app will automatically pull in events like 'lead form submission' and deal or opportunity stage change.
Below we see an example of a visitor’s journey with events like ad touch, conversion, or type in traffic included. At the moment the specific dates are not as important as the concepts.
- Attribution App uses a multi-touch Attribution model which considers all your ad channels when providing credit for a conversion.
- Attribution App installs a cookie in your visitor's browser to see each specific event along the visitor's journey.
- There are multiple types of events like ad channel touch, conversion, and direct traffic, you also can create your own custom events (and fire them) AND/OR integrate with conversion tools to track events automatically
Looking back vs looking forward:
When you compare Attribution app to your other analytics tools you may find major differences for a few reasons. One of the most common is due to the ‘multi-touch’ feature of Attribution app where it considers all the ad channel touches along a visitor’s journey rather than only one.
Another major difference is based on how Attribution uses visits and events to calculate revenue, conversions, and return on ad spend. While most other analytics tools will use an arbitrary look back window based around the conversion event, Attribution uses a precise and proprietary look forward window instead.
The look back window
Below is an example of a common look-back window, the opposite of what Attribution App uses, here the conversion event occurred on February 15th and a look-back window was generated to see which ads were responsible for this conversion.
With this method only the ad channel that was touched on January 15th would receive credit for this conversion, whilst the touch on December 28th would be ignored since it is outside the look-back window range. We find this to be an arbitrary look at your return on ad spend, instead we deploy a look forward window.
Look Forward Window
Attribution app deploys a look forward window when the visitor records their first event. The look forward window will track the visitor along their journey as they touch additional ad channels or continue to record events like conversion and sign up.
Because Attribution App tracks the visitor’s events instead of generating an arbitrary look back window, you will be able to see exactly which visitors completed certain events at specific times giving you high fidelity on your conversion, ad touches and revenue data.
Direct vs indirect traffic:
There are two primary types of traffic to be aware of, one type comes from your ad channels whilst the others are considered ‘direct.’ Knowing the difference will help you understand which model customizations to make.
In Attribution App direct traffic is considered any of the following
- Type-In Traffic - is when someone simply types in your URL from memory and visits your site. It is truly the only real "direct" way for them to get there.
- Bookmark Traffic -If a user has your site bookmarked, no referrer information will be provided in the request.
- Email Traffic - Emails with links to your site will not be tracked back to the email if you do not include tracking parameters. You should always provide tracking parameters in the links you include in your emails.
- Document Link Traffic - If you have links to your site embedded in a document such as Microsoft Word, or a PDF, there will be no referrer and therefore the visit will be segmented into Direct Traffic unless you include tracking parameters.
- Secure to Non-Secure Site Linking - If your site is insecure (http) and the visit is from a secure site (https) the referrer will not be passed from the secure site and the visit will look like Direct Traffic.
- Links from Native Mobile Apps (like Facebook) - If a visitor clicks a web link in a mobile app, such as the news feed in the Facebook native mobile app, the app will not pass along referrer information and it will appear that it is Direct Traffic.
Hopefully you feel comfortable with the look forward & multi-touch attribution concepts, however if you require any assistance please feel free to reach out to our support team at Support@attributionapp.com
There are a few customizations you can add to your multi-touch attribution model to make it fit your business better.
The first customization to consider is how you will attribute revenue and conversions based on the timing of events along the visitor journey.
Each customization will generate a new look forward window for specific reasons, the look forward window is where all the attribution will be derived from for the future conversion.
There are four customizations to consider:
- All Direct Traffic - This means any event (ad touch or direct) recorded after a conversion will generate a new look forward window.
In the example above three conversions would be recorded as a new look forward window was generated every time an ad touch or direct traffic / type in event was recorded within the date range. The return on ad spend would be derived from the touches that occur in the ‘look forward’ window prior to the conversion. This is often best for companies that want to consider 'direct' traffic as a credited event.
2. Include all traffic until a cut off event - This means any direct traffic or type in traffic that occurs after your selected ‘cut off’ event will not generate a new look forward window meaning it will also not be attributed.
This is often used by SaaS companies who only want to attribute to their ad channels after the customer has signed up or subscribed, since visitors will often visit the site via type in to log in and interact with the software.
Notice above that prior to the ‘cut off event’ Attribution app will attribute conversions to direct and type in traffic prior cutoff event. When you choose ‘all traffic included until a cut off event’ Attribution app will only generate a new look forward window for an ad touch event and ignore direct traffic after the cut off event.
3. Exclude All Traffic - This means Attribution app will never consider direct or type in traffic in your Attribution model.
In the example above Attribution will never generate a new look forward window for direct traffic or type in traffic.
4. Exclude All Traffic after a cut off event - This means Attribution app will ignore all traffic after a specific cut off event is registered.
In the example above only two look forward windows would be generated, after the cutoff event indicated by the blue triangle is registered along the visitor’s journey all attribution will be credited to the events prior.
This is a good customization for a company where a visitor may sign up for a tool or app and visit often. In this scenario you could choose to ignore all additional traffic after events like ‘sign up’ or ‘subscribe’ this would allow you to only consider attribution for the events that caused the initial sign up.
The value of a visit, putting it all together
Now you that you are aware of how Attribution app uses and generates the look forward window we can consider the value placed on each specific event in the look forward window.
We will begin with the simplest example first, the linear model.
- Linear Model - In the linear model each visit is provided the same credit for the conversion.
Now the example above can be a bit confusing at first glance, but let’s walk through it left to right.
When the first look forward window is generated attribution for the conversion is applied to the ad channels equally, three touches, 33% each for a total of 100% or one full conversion.
When our second look forward window is generated we see a 50% split between the two visits prior to the cut off event. This is due to our settings, because we have selected to exclude all traffic after our conversion event none of the visits after the blue triangle will be considered in the model and all the revenues and conversion events will be attributed to the touches prior to the cut off.
2. First touch - All credit for any conversion or revenue events will be applied to the first touch in the attribution model.
Here the first touch when the look forward window is generated will receive all the credit for the conversion.
3. Last Touch - All credit for any conversion or revenue events will be applied to the last touch prior to conversion.
In the example above, all credit is given to the touch prior to conversion, the direct opposite of first-touch.
4. Position-Based - This model allows you to set a value for the position of a touch in the visitor journey. You can adjust this value by going to Settings > Project Settings > Position-based Attribution ratios. You are welcome to choose a ratio for the first and last touch, the middle will be split amongst the remainder.
For the example below we will use a 40:20:40 ratio or a ‘U’ shape.
You may notice that in the second look forward window we split the middle 20% into two 10% credit touches. It is important to note that the middle ratio will always be split amongst the touches in a linear fashion. If our ratio was
5. Time Decay - This model will provide greater credit to the touches that occurred closest to the conversion using a 7 day basis.
If the sales cycle involves only a short consideration phase, the Time Decay model may be appropriate. This model is based on the concept of exponential decay and most heavily credits the touch points that occurred nearest to the time of conversion. The Time Decay model has a default half-life of 7 days, meaning that a touchpoint occurring 7 days prior to a conversion will receive 1/2 the credit of a touchpoint that occurs on the day of conversion. Similarly, a touchpoint occurring 14 days prior will receive 1/4 the credit of a day-of-conversion touchpoint.
If you run one-day or two-day promotion campaigns, you may wish to give more credit to interactions during the days of the promotion. In this case, interactions that occurred one week before have only a small value as compared to touch points near the conversion.
Don’t get distracted by the values here, the important concept to understand is that the touches closest to conversion will get more credit.
If anything in here is unclear please always remember that you are welcome to reach out to Support@attributionapp.com at any time.