A beginner’s guide to marketing attribution in an omnichannel world

Attaining the attention and loyalty of today’s always-on, distracted customer is becoming a significant challenge for businesses as they compete in a highly mobile and digitised marketplace. According to Aberdeen Group Inc., companies that have a strong omnichannel customer engagement strategy in place retain nine in 10 of their customers (89%), compared to only a third for companies that do not (33%).

It has been over a decade since marketing attribution capabilities first came onto the market to measure the performance of marketing and media against specific success criteria. Yet – with the volume of performance and audience data growing by the day – marketers are still confused as to how attribution fits into their overall marketing strategy, and the benefits it provides for optimising budgets and creating relevant experiences that drive meaningful business results.

What is marketing attribution?

To drive brand engagement, leads, sales, conversions and other desired business outcomes, most businesses will implement a variety of marketing and advertising tactics (in the form of paid, owned or earned media campaigns) over a period of time. In simple terms, the objective of marketing attribution is to accurately assign credit to each and every touchpoint along the customer journey that is proportional to their influence over the purchase or other business goal.

By understanding the influence that a particular touchpoint, or a combination of touchpoints, has on the customer’s decision, marketers can adjust their future campaigns as necessary.

However, today’s journey to purchase is becoming increasingly complex, with a tangled web of interactions occurring across both addressable and non-addressable channels, and increasingly, multiple devices.

Addressable channels are those that are trackable at a user level, and are usually digital interactions between a brand and its prospects or customers. Non-addressable channels, on the other hand, normally include offline interactions, and cannot be linked back to an individual consumer. For example, how can a brand know if a customer has been exposed to its radio ad before making a decision to purchase?

Marketers are therefore looking for a comprehensive way of measuring the impact of their efforts, regardless of the touchpoint type. But it can be hard to know where to start.

Approaches to marketing attribution

Multiple attribution approaches exist and which one a company decides to use depends on its goals, business requirements and level of marketing sophistication. For example, a brand that wants to understand the impact of each channel in driving total sales on a quarterly basis needs a different attribution strategy from a brand that wants to understand yesterday’s top performing keyword in driving online sales.

While attribution can help marketers garner meaningful insight and take smart action, the questions that various attribution approaches will be able to answer will differ significantly.

Marketing mix modelling (MMM)

Marketing mix modelling, also known as top-down attribution, involves amassing and analysing a large amount of historical, summary level data with the aim of providing broad recommendations on how to best allocate budgets across different channels and sub-channels to optimise performance. This approach is also useful for understanding how external factors, such as weather, the economy, etc., impact overall performance.

The lack of granularity and infrequency of reporting dramatically impact the marketer’s ability to analyse performance and optimise for future campaigns.

The primary benefit of top-down attribution is that it can be used across online and offline channels, providing marketers with a holistic view of their entire marketing mix.  

The difficulty with this method however, is that the datasets collected often come in a variety of formats, requiring additional manpower to manually normalise the data, and the help of data scientists to make sense of it. Furthermore, while the insights and recommendations derived from top-down attribution can answer key channel/sub-channel allocation questions, the lack of granularity and infrequency of reporting dramatically impact the marketer’s ability to analyse performance and optimise for future campaigns.

TV attribution

By using a bottom-up methodology to establish a baseline for digital marketing performance, TV attribution allows marketers to compare results with and without TV impressions to determine their impact on digital responses.

This is a valuable way for marketers to understand how TV drives digital engagement with a brand, such as website visits, search queries and even conversions. Performance can be evaluated by network, programme, spot length, creative and other dimensions to inform tactical optimisations.

Unfortunately, some vendors attempt to link TV impressions to specific users, based on their online behaviour once a TV spot has been broadcast. By doing this, vendors are effectively adding a TV touchpoint to the conversion journey, and making a statistical assumption about an individual’s exposure to a TV impression. This can compromise the integrity of the data and the accuracy of the insights marketers need to inform effective optimisations within TV and digital channels.  

Multi-touch attribution (MTA)

Multi-touch attribution, also known as bottom-up, focuses on addressable channels and can analyse both online and offline touchpoints. Unlike top-down attribution models, MTA makes use of a unique identifier to track users across channels and devices, providing a deeper understanding of the effectiveness of different marketing tactics along the customer journey, at a highly granular level.

The challenge with multi-touch attribution is that not all marketing tactics are addressable, which means that credit is only assigned to the trackable touchpoints in a consumer’s journey.

The benefit of this approach is that the models can be rebuilt daily, enabling marketers to continually optimise in-flight campaigns based on the most accurate, up-to-date-metrics, rather than waiting for post-campaign analysis.

Moreover, rather than tracking total sales like MMM, MTA allows for multiple types of online and offline KPIs, ranging from video views and form completes to in-store sales and repeat purchases.

The challenge with multi-touch attribution is that not all marketing tactics are addressable, which means that credit is only assigned to the trackable touchpoints in a consumer’s journey. Additionally, the accuracy levels of the various bottom-up approaches available can differ dramatically.

Multi-touch marketing attribution can be broken down into three primary subsets: 

• Rules-based attribution uses subjective rules to allocate credit to each marketing touchpoint. Rules-based approaches include first click, last click, even weighting and U-shaped, and are generally more simplistic, only allowing for basic levels of optimisation.
• Path-based attribution uses algorithms to compare customer journeys across media touchpoints to isolate the fractional impact that a particular touchpoint has on a specific KPI. Although this approach is considered more sophisticated than rules-based methods, it requires individuals to experience similar conversion paths to accurately assign fractional credit to a given touchpoint. Since you can’t force a consumer to take a pre-defined journey, accuracy issues arise when attempting to allocate fractional credit at granular dimensions (placement, keyword, creative, etc.). Therefore, this approach is more effective for measuring the impact of broader marketing channels.
• Multi-dimensional attribution tracks a user’s path across addressable marketing touchpoints using a unique user ID. Rather than comparing user journeys however, it models each touchpoint independently, and in combination, to understand how exposure impacts consumer behaviour. This approach enables marketers to optimise at a high level of granularity, and the models can be rebuilt on a daily basis to produce more accurate, up-to-date measurement insights and optimisation recommendations. However, because it is limited to addressable channels only, it is best implemented alongside TV and/or top-down attribution to ensure offline marketing tactics receive accurate attribution credit.

Marketers are rapidly beginning to realise that to successfully engage consumers they must employ in-depth analysis of the entire journey to conversion both online and offline, as well as across devices. Marketing mix modelling, TV attribution and multi-touch attribution each provide different but vital information for improving marketing effectiveness. While understanding the nuances of each approach is important, selecting the right solution for your organisation comes down to your goals and objectives, and the type of insight you need to drive meaningful results for your business.