Sign in

Create a profile

Get the most from McKinsey.com:

  • Choose from 40 newsletters and alerts
  • Save, download, and print unlimited articles
  • See personalized content on our app
Our Insights > Multi Channel Attribution – A Guide for Business Owners

Multi Channel Attribution – A Guide for Business Owners

10 July 2022

With so many marketing channels driving traffic to your business, how are you supposed to know which one is responsible for pulling in customers? Was it that blog article you wrote six months ago, or the Facebook article you posted last week?

To help you analyse which channel is converting into sales, you need to use a multi channel attribution model.

What are multi channel attribution models?

Attribution models are frameworks that help you determine the value and conversion rate of each touch point in your marketing channel. These touch points are the points of engagement that provide information about your business to the consumer.

The six most popularly used multi channel attribution models are: 

  • First interaction model
  • Last interaction model
  • Last non-direct click model
  • Linear model
  • Time-decay model
  • Position-based model

Using a multi channel attribution model can help you analyse the performance of your marketing channels, run tests and gather data to better optimise your conversion rates. Let’s take a look at each of the models in more detail.

First interaction model

With the first interaction model (also known as the first-click model), the first touch-point that the consumer interacts with is the one that receives all the credit for bringing in the customer. Any other touch-points that could have been interacted with before the purchase are ignored.

For example, if the customer finds your company through Twitter, this channel will get all the credit regardless of whether they viewed your paid ad two weeks later and visited your site.

Advantages:

  • Simple to implement and straightforward to understand.
  • Effective if you often get conversions quickly after the first touch-point.

Disadvantage:

  • You miss out on valuable information and the impact of channels later in the cycle.

When to use this model

If your industry has a short buying cycle or you need to bring in a large quantity of top-of-funnel customers.

Last interaction model

Similar to the first interaction model, in the last interaction model (also known as the last-click model), the last touch-point that the customer interacted with is the one that receives all the credit for bringing in the lead. Any other touch-points that led up to the purchase are ignored. This point is often just before the conversion.

For example, the customer might have interacted with the company multiple times but the point that will be counted is the moment they go to your website directly and make a purchase. This model is often the default model used by platforms including Google Analytics.

Advantages:

  • Simple to implement and straightforward to understand.
  • More accurate, as the last interaction before the conversion is certain.

Disadvantage:

  • You miss out on valuable information on the impact of channels earlier in the cycle.

When to use this model

The last interaction model is great if your industry has a short buying cycle and there aren’t many touch-points before your conversions. You can also benefit from this if you have a sales funnel that is wide at the top and narrow at the bottom.

Last non-direct click model

With this model, the interaction that gets the credit is the last interaction before the direct search or the paid ad click. This helps to eliminate any direct traffic before the conversion.

Advantage:

  • Eliminates direct traffic where your site is known and has been entered directly via the URL or a bookmarked link.

Disadvantage:

  • It assigns 100% of the value to a single interaction and misses out key information.

When to use this model

If you want a model that’s similar to the last click model and has a lot of direct traffic links for you to analyse. It can be hard to determine how customers that enter via direct traffic knew about you to begin with.

Even Google Analytics and other platforms have trouble identifying the source of direct traffic. Traffic will sometimes be marked as direct if the platform is unable to pinpoint where the visitors came from.

Linear attribution model

With this model, you split the credit for the conversion equally between all the interactions and touch-points a customer has had with your business.

For example, if a customer bought a £200 digital course through a paid ad and you have 5 touch-points, each would get 20% credit and are valued at £40.

Advantage:

  • A more balanced way of looking at your marketing channels in comparison to other models with a single point of focus.

Disadvantages:

  • Too simplistic and often inaccurate because some touch-points are more effective than others.
  • Hard to perform conversion rate optimisation with because all points are assumed to be valuable.

When to use this model

If you want a model that is simple and easy to explain to your clients, this will demonstrate the value of each channel.

Time-decay model

This is similar to the linear attribution model as it spreads the value across multiple channels but instead focuses on when the touch-point occurred. The point closest to the time of purchase is more valuable than the point furthest away.

Advantage:

  • A more balanced way of looking at your marketing channels in comparison to other models with a single point of focus.

Disadvantage:

  • Too simplistic and often inaccurate because some top-of-the-funnel marketing techniques may be very effective.

When to use this model

This model is best if your industry has a long sales cycle such as those with expensive B2B purchases.

Position-based model

This model splits the credit for a customer purchase between the first and last interaction. This has also given it the name “U shaped” attribution. With this model, 40% is given to each of these points and 20% is spread out between the rest.

Advantage:

  • A balanced way of viewing your marketing channels whilst giving more weight to the most important interactions.

Disadvantage:

  • Some middle-level marketing techniques may be points that have contributed the most.

When to use this model

If you want a focused model that still acknowledges the importance of all the touch-points in your marketing strategy.

The bottom line

Each of these multi channel attribution models can be used to determine the ROI for all of the marketing channels you have in place. It is worth noting that there is no “best” model. They all can be used in different circumstances and all come with their own benefits and drawbacks.

The most common difficulty in trying to decide which model to use is that they all have their own flaws, and many focus too much on a single point in the cycle. You will need to decide which point is the most important factor in your strategy and which touch-points you are willing to forgo the value of.

Increase your conversion rates

If you are looking to increase your conversion rates and maximise your revenue, it is important that you consider your multi channel attribution models and set your marketing channels up correctly.

As a digital consultancy with 15 years of marketing experience and a focus on conversion rate optimisation and paid ads, we have helped many companies achieve their goals and raise their revenue.

If you need help with setting up your marketing channels in a way that maximises your returns, get in contact with us today.