Attribution models in web analytics

In today’s competitive markets, a well thought-out, multichannel marketing plan is often the cornerstone of success in online business. For online shops, the correct marketing strategy is especially decisive in helping your business get ahead of the competition in e-commerce. When compiling this plan, every channel has its place. The customer journey is normally a lengthy one, with many different points of contact where the customer meets the product or brand. Some of these points of contact are more important for the eventual purchasing decision, while some are less important. So how do you work out which are the most important for your company?

Attribution marketing can help you analyse the different points of contact and measure the value of individual advertising strategies and channels. The right model for you should ultimately help you spread your marketing budget most optimally across different channels and maximise your ROI (Return on Investment).

The problem with the marketing plan

Most companies incorporate a varied marketing plan, consisting of e-mail marketing, SEA, SEO, affiliate links, and social media marketing. Which of these five main marketing strategies you focus most of your effort and budget on will depend on the target market you’re aiming at, though it’s worth bearing in mind that SEA (with its paid ad displays in search engine results pages) tends to be the biggest cost factor. But one thing that’s clear is that, as a marketer, you should aim to get as much as you can out of each channel. Your Return on Invested Capital (ROIC) should also be as high as possible.

In order to identify potential for optimisation in the marketing process, companies should familiarise themselves with the route that users take from first contact to eventually purchasing the product. This route – known as the customer journey – can be broken up into different points of contact, known as ‘touchpoints’, at which the customer interacts with the brand in some way. An example of a customer journey with several touchpoints could be as follows:

  • A sponsored post of a product on a social network catches a user’s eye
  • Through retargeting, an ad is later displayed to the user
  • Soon after, the user searches for the provider and product on Google
  • The user finds their way to the shop, buys the product and becomes a customer

Once a user has registered at the online shop and placed an order, they are considered ‘converted’ – meaning that they have become a customer. The customer journey described above is just one example of a common path through multiple touchpoints that can result in a sale being made and a user being converted. The three points of contact in this example were social media marketing, retargeting (also known as display advertising), and SEO.

So which of these touchpoints was the most decisive for the purchasing decision in this example? And what percentage of overall revenue is each responsible for? Attribution marketing attempts to find models which help you to answer these questions.

What is attribution in marketing?

The term ‘attribution’ describes the appreciation of cause-effect relationships in real actions and operations. The word comes from psychology, but has found its way into every marketer’s handbook.

In the case illustrated above, the points of contact for our user on their customer journey are the possible causes and the conversion is the end effect. For example, if a user has four touchpoints via different channels, then you as a marketer can identify each touchpoint separately as the decisive factor in the attribution model and focus the marketing budget on that corresponding channel. But which is the right strategy for attribution marketing?

Multichannel tracking as a starting point

Web analytics offers numerous options to help you understand the success of marketing strategies and to optimise your own. The classic way is through isolated tracking of individual channels.

Multichannel tracking is the term for this process, and it involves marketers trying to imagine the big picture and take an overview of their complete marketing plan. This is achieved by including all the possible measures that could play a part in the customer’s route from first encounter to conversion, to try and plot as accurate a customer journey as possible to work from.  Thanks to the analytical tools available today, marketers aren’t restricted to analysing each channel performance individually. Instead, you can observe the effects of several channels simultaneously, before deciding to split your budget according to the results you discover. The decision-making factor here is the different attribution models, which can be used to evaluate the individual touchpoints of the customer journey and assign each touchpoint a role and a significance.

Attribution modeling in online marketing

Depending on the attribution model, marketers place different values on the various points of contact throughout the customer journey. These models help to check which channels are the most important and should be prioritised.

Each channel is given a monetary value relating to the usefulness and successfulness of its role in the customer journey. The balance across all advertising channels is then used to determine the split of the marketing budget. This is known as optimisation. Below are several common attribution models from current marketing practices.

  • First click attribution: this model values the first click as the most important point of contact for the conversion.
  • Last click attribution:here, the last click is considered the decisive moment for converting a user to a customer. This model is more appropriate for low-cost consumer goods, where little consultation is required to make a purchase. In the past, the last click attribution model was the most commonly used method. But today, with the increased variety of advertising mediums available, attributing 100% of a conversion to the final channel that leads to the purchase is considered outdated and inaccurate for many industries.  
  • Linear attribution: the linear attribution model splits the revenue attribution evenly across all channels. Every point of contact on the customer journey is considered equally important to the end result – a conversion.
  • Time decay attribution: the time decay model makes time the central factor in attributing values to different channels. The quicker a channel’s point of contact leads to a conversion, the higher its monetary value is.
  • Position based attribution: this attribution, which is also known as the bathtub model or the U model, values the first and last point of contact more highly than the others. The touchpoint that initially introduces the potential customer to the brand or product and the touchpoint that finally seals the purchase are assigned the highest monetary value. The other steps in between take an equal share of the remaining conversion value.

These are some commonly used models, but you can define your own that best fits your company or campaign goal. A custom attribution, as this would be known, could rank each individual channel differently, based on a variety of factors like position, time decay, or existing traffic. Further options include different attribution models for different customer groups (new customers vs. existing customers) or different product groups.

Summary

The great challenge when it comes to marketing attribution online is to find the right model for your company. There’s no one single solution that’s proven to lead to success for everyone. The most important thing is that you correctly interpret the customer journey model for your product or brand. Rather than relying on good luck or guesswork for this, there are dedicated service providers that specialise in attribution on hand to help. This dynamic attribution modeling is based on a mathematical analysis of web analytics data for a company. Once the attribution has been implemented, it becomes easier in future to identify the most successful channels of your marketing strategy and to push your optimisation further. It’s particularly important to pay attention to information on inter-channel dependency, to maximise the effectiveness of your ad campaigns and utilise your budget the best way possible