For marketers, determining the where and when of a purchase is easy. The harder task is defining the reason behind a purchase — that’s where marketing attribution steps in.
Customers are exposed to a brand an average of 36 times before finally making a purchase, often interacting with multiple marketing channels and campaigns along the way.
With so many variables impacting a sale, it can be difficult to determine which interaction, campaign, or channel was the prime contributor that led to that conversion — especially if you have data only about the purchase itself.
Marketing attribution gives marketers deeper visibility into which channels are most effective in driving conversions, which also helps maximize ROI. Attribution lets your team build a holistic picture of the entire customer path to purchase, from first contact through conversion.
This guide has all the information you need to start practicing marketing attribution, including an overview of the many marketing attribution models in use today and how your marketing team can implement them.
In this marketing attribution guide, you’ll learn:
- What is marketing attribution?
- Why is marketing attribution important?
- Marketing attribution models
- Marketing attribution strategies
- Get started with marketing attribution
- Frequently asked questions
What is marketing attribution?
Marketing attribution is the process of determining which interactions influence a customer to purchase from your brand — it lets you know what route a customer took to your products.
With this information, marketers can identify which campaigns or channels drive the most conversions. If you’ve ever been asked how you heard about an online store, you’ve been part of marketing attribution.
Using attribution in marketing also helps marketing teams determine how to target campaigns to reach more high-converting customers or to retarget existing customers for additional purchases.
Marketing attribution helps align marketing and sales data at all stages of the sales funnel — from initial lead qualification through to conversion.
The impact of marketing and sales automation
When marketing and sales teams can coordinate their activities effectively, they enjoy 36% higher conversion rates and 38% higher sales rates. A marketing campaign can also benefit from attribution analysis.
The complex, multifaceted nature of modern marketing and sales strategies makes marketing attribution a challenging and transformative venture.
Today's customers may engage with brands:
- Digitally, on a range of devices and platforms
- In-person, at retail stores, trade shows, or events
This makes it crucial for marketers to accurately map the customer journey that leads to a conversion.
But the sheer volume of marketing and sales data means that manually identifying the most valuable touchpoints during the journey is not feasible. Marketers need tools for automating the data collection and analytics process.
To place marketing attribution in context, consider a customer who purchases a phone from a retailer that has both an online and brick-and-mortar presence. The customer ultimately makes a purchase online, but they visited a store to look at devices in-person prior to that event.
This customer was also influenced by multiple marketing channels such as:
- Online ads
- In-store marketing materials
- Interactions with salespeople
If that retailer knows only which device the customer used to make the purchase and that they did so online, they have little insight into the touchpoints that were most important in driving the customer to convert.
Why is marketing attribution important?
Marketing attribution offers a range of benefits to marketing teams:
- Future personalization opportunities. When you understand the factors behind each conversion, you are in a better position to personalize your marketing and sales efforts for future purchases.
- Higher ROI. It's critical to know which channels deliver the highest conversion rates if you want to maximize the ROI of your marketing efforts. Attribution reveals that information.
- Product improvement. Understanding how customers approach your brand and its offerings can give you insight into the features or value that customers are looking for. With that data, you can improve your products so that they align with customer needs.
- Justify marketing value. Marketing attribution helps to justify marketing expenditures to win buy-in from stakeholders. For example, if your CEO asks why you are spending $100,000 on billboard ads when most of your purchases are made online, marketing attribution would allow you to respond with data that proves billboard advertising is effective.
Marketing attribution models
There are a variety of approaches to marketing attribution, each oriented around a different type of analytics model and approach to weighing the various data points that influence customer behavior.
At a high level, these models can be broken into two categories:
- Single-source models, which are simpler and focus on identifying just one touchpoint that is most influential in conversion.
- Multi-source models, which are more complex and weigh the relative influence of multiple touchpoints.
Now let’s look at the various attribution methods available within these categories:
First-touch marketing attribution
This is perhaps the simplest marketing attribution model. It assigns complete credit for conversion to the marketing effort that first brings a customer in.
This model is easy to understand and implement. However, it measures only top-of-funnel marketing efforts and provides no real visibility into important touchpoints that happen later in the sales funnel.
Last-touch marketing attribution
This model, which is also easy to understand and implement, inverts first-touch attribution by focusing on the final touchpoint that leads to a sale.
The drawback to this model is that it tracks only bottom-of-funnel marketing activity. It is not useful for measuring the channels that bring customers to your brand in the first place.
Linear marketing attribution
Linear attribution is a simple multi-source attribution model that aims to identify all key touchpoints that were relevant in driving a purchase. Once those touchpoints are identified, equal weight is assigned to each one.
This model is relatively simple to implement because it doesn’t give different weights to different relevant touchpoints. However, this is also a weakness — if some interactions were more important than others, this model won't reveal the most influential ones. It only separates totally irrelevant touchpoints from those that were important in some way.
Position based marketing attribution
This model — also called “U-shaped attribution” — assigns 40% of the credit for conversion to the first touchpoint and another 40% to the touchpoint that immediately precedes a conversion. The remaining 20% of credit is split between other touchpoints.
This is a more sophisticated model that helps marketers interpret the relative influences of different touchpoints within a complex marketing and sales process. However, it is somewhat arbitrary in that it assumes that the first touch and the pre-conversion touch are the most important, which is not always the case.
W-shaped marketing attribution
Under a W-shaped attribution model, the first touch, the lead creation touch, and the opportunity creation touch each receive 30% of the credit. The remaining credit is distributed to other relevant touchpoints.
This model offers the advantage of highlighting three of the touchpoints that are often, though not always, the most influential on customers. Its main drawback is that it affords little weight to post-opportunity creation touches.
Time decay marketing attribution
This model assigns the greatest credit to touches that occur closest to conversion. It offers a multi-source model that is easy to calculate but is still able to support a large number of touches.
The downside to time-decay attribution is that it may undervalue lead generation and other top-of-funnel marketing activities because they typically happen relatively far away from conversion.
Choosing a marketing attribution model
Think of the marketing attribution models above as guidelines, not prescriptions. You can modify them to fit your team's needs, using your own custom algorithms.
The best model for you will depend on how many touches you typically have before a conversion, and the way those touches are distributed across the sales funnel.
If your customers engage only a handful of times before converting, a simple, single-source model may work well for you. If you invest most heavily in end-of-funnel marketing activities, choose a model that allows you to measure multiple touches that occur late in the funnel.
Variables to consider when choosing an attribution model
The key things to weigh up when selecting between marketing attribution models are:
- The sales cycles used
- Whether offline or online channels are dominant
- What software you’re using.
Your knowledge of the typical sales cycle in your company impacts the decision because each buyer journey can be unique. Certain industries can have up to double the number of typical touch points as others, so if this sounds like your sector, a single-touch attribution model may be less than ideal.
If your sales cycle lasts longer than 90 days, then a first-touch attribution model may not pick up your top-of-the-funnel campaigns at all.
Firms with multiple touch points of equal weighting might find a linear attribution model is best for them, since this places weight equally on each. If budgets vary considerably between touch points — for instance, if you host conferences – then a model that recognises the importance of decision-making may be better. This will often be the W-shaped attribution model.
How much offline-online interaction your business has will also make a difference. Those with more offline presence may need to rely on workarounds where there are tracking gaps, such as campaign-specific codes that could link an in-store purchase back to a TV commercial.
Software also has an impact. If you’re using Google Analytics, then the seven-day half life often doesn’t leave room to accommodate a time-decay analysis.