Driving Revenue with Marketing Attribution

Why Marketing Attribution?

When a business plans marketing initiatives, leaders look to the prior campaigns, marketing results, and ROI.  Which channels were most cost-effective? Which landing page drove the most leads? What customer paths provided the largest transaction value. However, if leads, conversions, and revenue cannot be attributed to specific channels, ads, and marketing elements, how do brands know what’s working?

Marketing attribution provides a framework for measuring touchpoints throughout a customer journey.  Multi-touch attribution (MTA) helps identify the true lead source, the marketing elements driving conversions, and calculates the value of each marketing element towards a sale. Combined with a brand’s business model, customer profile, and marketing mix, MTA uncovers the value of each customer journey, eliminates waste, and increases revenue.

Marketing attribution ties marketing elements to realized sales, closing the marketing-sales loop.

Analyzing Aggregate Journeys

With high-volume businesses, often aggregate data alone surfaces actionable insights. Data can show a clear differentiation in not only volume, but in behavior such as lead conversions by channel and segment. Aggregate data also reveals where to begin drilling down into the customer journey and the questions needing resolution. 

In the Higher Education examples below, the difference between submitted applications and recognized revenue is revealed through full-path attribution. 

Total Unique Page Paths by Source

Each channel produces a set of unique page views which can be filtered between total views, total views producing conversions, and total views creating revenue. Examining the difference between the data sets reveals which pages and paths are contributing to lead conversion, and which ones are producing revenue, and which fail to impact customer behavior positively. 

 

Viewing unique paths by source and revenue provides an accurate efficiency rate for each channel.  In the case above, Google Ads had an efficiency less than 5%; organic traffic produced revenue on 8% of visits. Comparing total path visits with revenue-producing visits uncovers the true value of each channel. In this example, the total visits appeared to show strong paid media results. However, filtering for revenue exposes organic traffic as far more valuable.  

Conversion Page Visits

Many brands, especially B2B, use mini-conversions to capture leads to nurture toward sales. These can be a white paper download, application for an MBA program, or request for a product demonstration. In addition to providing a method to connect with leads, the conversion creates a data point for analysis and optimization.

By comparing all conversion page visits with just those that produce revenue, the actual value of each landing page is calculated. Filtering further by organic, paid media, specific campaigns and even keywords provides robust optimization data. With multi-dimensional filtering and full-path attribution, businesses can determine:

  • Which pages are producing organic leads 
  • Which paid media campaigns are producing leads 
  • Which landing pages are creating revenue
  • Which campaigns are wasting resources
  • Which spends should be reallocated to performing ads
  • Which landing pages should be optimized to raise Quality Score 

With full-path attribution, brands can determine what is creating value, and where efficiencies can be gained by reallocating spends, remapping landing pages, and revising keyword strategies. Because user-event data and metadata is captured for every touchpoint, the opportunities for increasing revenue and reducing waste are exponential. 

Analyzing Individual Journeys

Reviewing aggregate data answers key performance questions, and provides starting points for diving down into individual journeys. By reviewing specific paths, marketers can not only uncover the specific touchpoint combinations that drive revenue, but analyze how each channel is delivering value.

Path Analysis

Analyzing specific paths by cohorts and diving into individual journeys, attribution data reveals the discrete components that are impacting performance by channel, conversion, and page.

Attainable Attribution Goals

To determine the credit each touchpoint receives for a sale, MTA requires capturing an accurate full-path record of all customer journeys. Marketers can then analyze customer journeys against different model views in aggregate and individual journeys, identifying trends, outliers, and the unanticipated. In short, they can calculate the value of each marketing element to revenue creation. 

After this full view of customer journeys is achieved, brands can pursue a number of critical revenue-driving goals:

  • Identify the most impactful marketing component

  • Allocate spends across marketing elements by impact
  • Justify expenses across marketing and sales initiatives 

  • Close the marketing to sales revenue loop

  • Integrate data into a CRM for a 360 view of the customer

  • Navigate decisions involving complex customer journeys across a marketing mix

Understanding the customer journey—in aggregate and in specific individual journeys—opens up opportunities for reducing costs and maximizing revenue. Once unfiltered and complete journeys are captured and visualized, the data can be sorted and transformed by key business dimensions. For example, queries can determine shortest journey, most profitable journey path, largest transaction value, lowest CAC, obstacles to conversion, and source of greatest LTV

Why Isn’t Every Business Using MTA?

While multi-touch attribution is more complex than some traditional measurement, MTA is well within the reach of small businesses. Similarly, some brands might believe that legacy platforms are an insurmountable obstacle.  However, they are not. In fact, some modern MTA platforms capture user-event data completely independent of ad platform APIs and can integrate with numerous CRMs and data visualization tools. 

The most significant barrier to attribution adoption is not technology but basic misunderstandings of how MTA best functions.  Because of the multiplicity of approaches and collateral online, many practitioners begin with unclear expectations of what different MTA modeling requires and achieves.

In recent years, about 75% of businesses report either using or planning to implement MTA; in fact, many make more than one adoption attempt (GoogleBCGGartner). While the challenges span unique contexts (and solutions), one thing is clear—businesses suffer from confusion in the pursuit of a measurement strategy. 

A Framework for Marketing Attribution

MTA begins with a capture of user-event data across all converting customer journeys.  This single source of data can be transformed, filtered, and sorted to answer nearly limitless questions about customer behavior across brand touchpoints. The queries can be pushed into dashboard visualizations, directly into the CRM, or natively pulled with SQL queries.

Data Capture

The core data + metadata can be captured through multiple techniques, including platform APIs, first-party cookies,  or javascript. However, the key to data usability is the capture of user-event data and the metadata that identifies all parameters and dimensions that contribute to business KPIs. For example, a user ID (connected anon ID and conversion ID) and location (e.g., page, screen, source) is connected with meaningful metadata (e.g., channel, campaign, ad creative, page URL, time, duration, etc.). The dataset is captured for all touchpoints leading to a conversion and provides the dimensions used in queries. 

Data Transformation and Queries

Once user-event data and metadata is connected in a single source of record, the resulting data store can be mined through sorts, filters, and queries. As a result of the ability to transform and visualize journey data by specific dimensions, analysts can quickly calculate and compare KPIs. For example, analysts and marketers can:

  • Calculate ROMI by customer segment, specific path combinations, and compare CAC and LTV by customer paths.
  • Analyze sales cycles by duration, transaction value, and sequencing correlations with value.
  • Compare marketing element performance in each funnel phase, identify anomalies and obstacles prohibiting key transitions through the journey.
  • Apply MTA models to compare journeys with differently weighted touchpoints to predict returns and identify optimizations.

Choosing a Solution

When considering an MTA solution, brands should focus on the key questions relevant to their marketing mix, customers, and sales cycle. Which questions help the business understand each key journey point (e.g., first visit, conversion to lead, conversion to customer)? Which KPIs and dimension combinations are important (or unique) to the business? Is a  large website with varied organic entry points a substantial portion of the mix? Is paid media a large percentage of the acquisition strategy? What are the key questions that an attribution solution should answer for the brand?  Consider these when reviewing solutions.

Brands should remember that specific MTA models are forms applied to uncover patterns that enlighten an understanding of the customer and drive profitable decisions.  Models are overfitted when they stop being contextual tools and become templates universally applied. Arcalea’s Galileo attribution is model agnostic; use any model you choose or compare actual individual and aggregate journey data. Optimize your marketing mix with real-time high-performing tactics and assets.

Most companies have specific questions about their marketing programs. As data is captured and the customer journeys become clearer, more questions will be discovered and answered. And if brand leaders and analysts have never seen a clear unfiltered view of all customer journeys, the discovery of hidden insights alone may be MTA’s biggest value. 

Many reasons drive adoption of MTA: complex customer journeys, sophisticated marketing mixes, failure of legacy measurement, etc. But if your attribution solution isn’t driving significant revenue and efficiency gains, you’re doing it wrong.

Leave a Reply