Quick definition

Attribution refers to the practice of assigning credit for conversions or revenue to marketing touchpoints in order to pinpoint the ones that are working best and allocate resources accordingly.

Key takeaways

● Attribution is mostly for marketing and for specific parts of the customer journey, but it can also be applied to other business functions.

● Different types of attribution models measure different marketing touchpoints. Last-touch attribution, for example, focuses on the final step that led to meeting a business goal.

● Adobe Analytics gives you access to much more specific attribution models than are available elsewhere. The evolution of attribution makes it possible to measure the business impact of actions taken across the entire customer journey.

Nate Smith is a group manager of product marketing for Adobe Analytics Cloud. He oversees strategy development, product launches, positioning and messaging, pricing and packaging, sales enablement, product requirements, and competitive analysis for several Adobe products. Nate has been with Adobe for almost 11 years, and before his time at Adobe, he had six years of experience as a marketing strategist.

What is attribution?

How does attribution divide credit among marketing touchpoints?

How often should you check your attribution model?

How specific can attribution models get?

How does an attribution model look and work?

Is it difficult to use an attribution model?

Is there a recommended attribution model to use?

Does attribution impact other parts of business besides marketing?

What are some challenges that come with attribution, and how do you combat those?

How will attribution be different in the future?

Q: What is attribution?

A: Put in the simplest terms, attribution is a way to assign credit to all your marketing touchpoints for revenue or conversion. It's a way to identify the marketing decisions that actually produced revenue. It also helps you allocate your marketing budget accordingly, dividing the larger sections of your budget to more successful touchpoints.

To better understand attribution theory, a great analogy to use would be a soccer match. Before a team scores a goal, the ball might hit several different feet, until one last kick secures the point. But each set of feet contributed to winning that goal. The same happens in marketing. Attribution pinpoints the marketing “kicks,” or choices, that led to winning the business goal, such as conversion or revenue. And marketers use attribution models to decide which types of kicks their companies should do more of.

Q: How does attribution divide credit among marketing touchpoints?

A: There are several different types of marketing attribution models. The three most common ones are first-touch attribution, last-touch attribution, and linear attribution. These models fall under a type of attribution model called a rules-based model. The model you choose for your company will reflect the different strengths of your marketing touchpoints.

First-touch attribution refers to the first marketing touchpoint that leads to meeting a business goal. Referring back to the soccer analogy, this is the first set of feet that the ball touches. The model will assign the most credit to this first marketing choice. This model is best to learn about ways to promote engagement.

Last-touch attribution refers to the final marketing touchpoint that happens right before meeting a business goal. The model will pay most attention and assign the most credit to the business deal closers. This model is best to learn about ways to achieve conversion.

Linear attribution refers to each marketing touchpoint that leads to meeting a business goal. Each kick, even the final kick, is assigned an equal amount of credit.

You can adjust rules-based attribution models to assign credit to whichever touchpoints you want to focus on. For example, you can use time-based attribution to assign credit to your most recent touchpoints.

Besides rules-based attribution, there’s also another category called algorithmic attribution. This type of attribution uses machine learning to determine where credit should be assigned. This approach is called the “best-fit model,” and some marketers prefer this model because it’s less work for them. However, it might be a good idea to apply some other types of models to compare to an algorithmic model.

Q: How often should you check your attribution model?

A: A great rule of thumb is to check your attribution model quarterly — at a minimum — and make appropriate adjustments. This regular maintenance lets you account for new types of marketing choices or changes to your marketing strategy.

Q: How specific can attribution models get?

A: Historically, attribution has been based on marketing channels, paid media being the most common and over-rotated channel. The problem with this, however, is that in focusing on paid media as your only channel, you’re missing out on deeper-funnel touchpoints.

Assigning a ton of credit to one broad touchpoint like paid media doesn’t help you improve something more specific, such as your email optimization. More useful technology would allow you to focus on deep-funnel specifics like campaigns and even search terms.

Adobe already offers this type of technology. With Adobe Analytics, you can assign algorithmic, best-fit models to things like email campaigns and determine which type of attribution to applies to other campaigns and channels.

Q: How does an attribution model look and work?

A: Typically, attribution data from an attribution model is shown as a table. In the left column will be the channels you’re pinpointing, and in the right column will be your metrics, such as revenue, item orders, or any other type of conversion metric you want to measure. The model will calculate the impact for each channel that you choose to include in your model. It will also show revenue percentage, contribution percentage, and other metrics you want to add. The same thing happens with an algorithmic model. If you’re using Adobe Analytics, you can even create visualizations of the data that you can show to stakeholders.

Q: Is it difficult to use an attribution model?

A: Attribution models are marketer-friendly. You do need some knowledge about data to make things easier, but for the most part, an attribution model is designed to help marketers understand how to allocate their budgets. The industry is pushing to make attribution more and more user-friendly. Adobe has already taken big steps to make attribution easy to use and learn.

A: Adobe’s main priority is to offer as many types of attribution models as possible. We don’t push one type of model, because we want you to choose the types of attribution that are perfect for your marketing strategy. There is no attribution model that provides every bit of information that you need. The types you choose will depend on what each user in your organization needs and each use-case.

For instance, digital analysts and channel marketers are tasked with answering questions about marketing, whatever those may be. Their use-case would be using attribution to garner data that would answer those types of questions. Operations’ use case for attribution is determining how to best optimize media spend. B2B companies’ use case would be account-based attribution, because of their longer sales cycles.

Q: Does attribution impact other parts of business besides marketing?

A: Attribution can be used for a lot more than just marketing touchpoints. For instance, you can use it for product development, including with digital products like mobile apps and websites. Attribution can help you figure out how you can refine your product.

You can also use attribution to improve customer service, especially with post-sale interactions. Attribution can help you understand what types of interactions work, and which don’t.

Even though attribution has mostly been used for ad spend, you can use attribution for any area of your business that you’re looking to improve, as long as it has a measurable result that you can trace to a choice.

Q: What are some challenges that come with attribution, and how do you combat those?

A: A big challenge with attribution is how easily it can be manipulated. Because people don’t want to lose their portions of the budget, they can run attribution models that purposely make their own work seem more impactful than it might be. To combat this, it’s important to have a level of governance over attribution modeling, to avoid different teams and departments fighting over control so that they don’t lose funding. You could assign this governance to a specific team and decide as a company what type of attribution models and lenses you will choose.

Q: How will attribution be different in the future?

A: To bring some context about how attribution has worked for now, the way to gather personal data about potential or current customers has been through third-party cookies. But companies like Verizon, Google, and Adobe are currently trying to find a new way to gather data for different personas, while still respecting customer privacy.

Besides focusing on privacy in future evolutions of attribution, the industry is also focusing on breaking down the siloed aspects of attribution. Instead of focusing on only one part of the customer journey at a time, attribution platforms are seeking to provide an experience that would help put the entire customer journey through the different lenses of attribution. Instead of just ad spend, the future of attribution will focus on the holistic customer experience. And the biggest piece of technology that’s going to help achieve this is AI and machine learning. The future of attribution points to highlighting specifics of the customer journey like cross-device ID, and even customer biases.

People also view