Home >Blog >Hey! Are you Paying Attention to Multi-Touch Attribution?
Creating Attribution Models

Hey! Are you Paying Attention to Multi-Touch Attribution?

Phillip Reinhardt

If 100 percent of your customers go through your online shopping cart, does that mean that the only advertising and marketing that’s working for your business is the single ad that sends people directly to the cart to make a purchase? Chances are, the answer is likely much more complex. Customer journeys today are like a pile of spaghetti, with prospects pinging around from site to site checking reviews, looking for alternatives and researching quality before ultimately making a decision to purchase. Since the journey to the shopping cart is so non-linear, how can marketers determine which tactics are actually working and which should be discontinued? Multi-touch attribution is one way of assigning a percentage of credit to a variety of different tactics or channels, but it can seem relatively complex at first blush. 

Attribution Approaches

Several different attribution approaches are commonly used by marketers, including:

  • Final touch, where only the last touchpoint receives all the credit for a sale.
  • First touch, where only the first touchpoint receives all the credit for a sale.
  • Marketing Mix Modeling, where complicated algorithms are used (often by massive CPG marketers) to drill down into data and assign a percentage of results to each channel.
  • Multi-touch Attribution, where advanced web analytics are leveraged to develop predictive models that observe consumer behavior in near-real time.

If sounds like this is getting pretty deep already, you’re right. Data scientists can have a field day with all of the information that’s available around marketing — and they do! It’s still difficult if not impossible to fully attribute each sales channel from mobile to email to SMS text to social. However, marketers continue to try and are able to make some educated guesses based on data where to funnel their advertising spend.

How Customers Reach The Shopping Cart

Types of Multi-Touch Attribution

Just as there are multiple ways to determine valuation of a channel, there are several ways you can calculate overall attribution: equal weighting, time decay or algorithmic. With equal weighting, assume that there were three steps in the customer journey such as email, web search and social retargeting ad. Each touchpoint would receive one-third of the value. With time decay, the most recent touchpoint receives a greater percentage of the credit for the sale. Algorithmic is the most complex, and relies on each individual customer’s journey and where each channel appears within the journey to assign credit for the sale. See examples of each as explained by TechCrunch, with detailed math showing the attribution for each channel.

Using Multi-Touch Attribution (Without Losing Your Mind)

Unfortunately, not every business has access to a team full of statisticians and data scientists just looking for some numbers to crunch. In the real world, marketers have to do the best they can, and that means we need to make things simpler and faster. How can you leverage the concepts behind MTA without losing your mind and drowning in details? Fortunately, we’ll break it down for you. Keep in mind, this doesn’t provide perfect results — but they may just be good enough to help you make business decisions.

In general, time decay is one of the better models to use. You can assign a certain percentage to each touchpoint. Let’s say you have a campaign for a particular item that includes a Google or Bing search ad, social posts, Facebook retargeting and a weekly email. You notice that nearly all of the conversions are happening due to the Facebook retargeting ad (last touch attribution) but you know that there are other things also pointing customers in the direction of a sale. With a time decay model, your attribution may look something like this:

[Insert table functionality is not working on WA platform, but I would put this info in a small table]

Google / Bing Search Ad  – 10%

Social Media Posts – 15%

Weekly Email – 25%

Facebook Retargeting – 50%

The reason for the staggered weight is that the touches that are earlier in the customer journey didn’t convert — so they shouldn’t be weighted as heavily as the items that allllmost converted the customer.

See how multi-touch attribution lets you distribute your marketing dollars more effectively? You can see that although the display ad on search engines is still an important part of your marketing mix, the weekly emails and retargeting ads are where you should really pour in some extra dollars. Contact PBJ Marketing today at 646-770-3271 or email us at makeithappen@pbjmarketing.com to see how to create an attribution model that works for your business.