Marketing and sales attribution: the science of the deal

Attribution in marketing and sales is how companies know how customers come into their sphere, and how people browsing the web go from anonymous shoppers to loyal customers. It wasn’t that long ago that companies, particularly B2B sales, had nothing to go on but guesses and rudimentary tracking. In today’s data-driven world, guesses no longer cut it. With automated marketing and sales attribution, you can get real, traceable facts and hard numbers to back up sales and marketing claims. Companies can go inside the buyer’s journey while it’s still in progress, make course corrections, feed the right content at the right time, and measure specific results. It’s a revelation for businesses, but attribution is miraculous for marketing and sales teams.

Marketing and sales attribution begins with rules and processes that define how to assign credit for various touch points in the customer journey. The question, “where did that customer come from?” is only the beginning. Naturally, the sales team says that their skill in the art of sales brought in a customer because, after all, they closed the deal. But the marketing team will often claim credit, too, citing that campaigns, content, promotion, and advertising originating from their efforts was what turned a prospect into a customer.

The truth is, they’re both right. Nutanix, a B2B multi-cloud platform and infrastructure company, tackled the problem of marketing and sales attribution by looking at the journey holistically. Using two different marketing and sales attribution models, the Nutanix RevOps team tracks touchpoints along a buyer’s journey—including marketing sources, channel data, and sales engagement.

“The results are only as good as the data”

Saumya Shah, marketing strategy and operations manager at Nutanix, understood that no attribution model could be successful without a clean data set. For any attribution model to work, Saumya said, defining and following processes is crucial. “If those processes are not followed, or if a certain behavior is not followed, the data is no good. The results are only as good as the data,” she told audiences in a recent Openprise webinar. “We have some processes defined and the correct behavior set for it.” For that, Saumya relied on Openprise, which also helps manage its marketing and sales attribution processes. “We use tools like People AI, Vital, PIA, and Salesforce to capture all that engagement and to house all the data,” she said. “And then we use Openprise for all the data preparation and the modeling and also to automate the processes.” Once they had processes for ensuring clean and complete data, they were ready to deploy their sales and marketing attribution models.

Winner takes it all: The source attribution model

One model Saumya and her team at Nutanix implemented is source attribution. This marketing and sales attribution model answers the question, “what triggered the opportunity creation—and what led to that creation?” They aren’t looking for how a contact got into a system but rather where the opportunity came from. “Any kind of attribution model—be it sourced, influenced, or whatever, it is very important to keep the notion of the sales in mind because all of our models run on top of our opportunities,” Saumya advised. That’s why building a healthy partnership between RevOps and sales is essential. Understanding how they do their processes is at the heart of their successes.

Nutanix’s attribution process under this model is based on contact roles added to opportunities. Whenever they create an opportunity, they add people to that opportunity as contact roles so that they can see engagements. “It could be a website campaign, a webinar, or a boot camp,” Saumya said. “We use Salesforce to track all our reports and to measure attribution.” And then they bring in Openprise again.

“Openprise looks up opportunity data and contact roles. Then, it looks up campaign member engagements across sales channels and marketing campaigns, and writes the data back into the opportunity object, so all of our reporting and data consumption still lives in Salesforce.”

From that data, Nutanix calculates the attribution split in terms of sales, channel, and marketing. “It’s a winner-take-all model,” Saumya noted.

But there are other winners as well—especially in RevOps. From running the Source Attribution model, they gather data on which types of campaigns bring in more opportunities. They’ve discovered that some campaigns are suitable for generating interest, while others perform better as opportunity creators, so they now have a firm idea of how each campaign will likely function.

Marketing and sales attribution for a multi-touch world: the influence model

It’s well-known in advertising that it takes many different impressions to create a customer. That adage also holds in RevOps, but fortunately, there are usually plenty of touches to go around. Companies often sell more than one product, and each line gets its own campaigns. Tracking and measuring a prospect’s engagement activities across multiple touches can get complicated. Companies need to know what prospects are doing after being added to an opportunity. They need to understand what influences their decisions to move forward with a purchase.

Influence attribution presents information that helps companies define their marketing tactics based on the consumer behavior within the buyer’s journey. Influence attribution depends on contact roles—without those all-important job titles, it’s impossible to see who’s influencing whom and to track engagement based on that. With a desire to become a more data-conscious marketing organization, Nutanix set out to build an influence attribution model that sliced the data differently.

“This influence model lives within Salesforce, so people don’t have to go outside of Salesforce to track it,” Saumya said. “All the inference detail records are within that object. And you can use opportunities, campaigns, contacts, and the influence the records created, and build all that into one holistic report in Salesforce.” Saumya added that this was for efficiency: using self-service to create reports gives salespeople the information they need faster and, at the same time, removes the burden of report creation from the analytics team.

Using the influence model, Nutanix gains an understanding of which content to feed to prospects. The model helps the marketing team plan targeted campaigns based on contact roles—with content that’s proven to work.

Why have two marketing and sales attribution models?

Nutanix settled on two main marketing and sales attribution models to get very different information from the same data. In the source attribution model, the team looks at a one-touch, winner-takes-all scenario—who or what brought the deal to the table. In the influence model, they can analyze the data and split the credit between all the factors that influenced the outcome.

“We do an even touch inference attribution model,” Saumya elaborated. “If a prospect is added to a webinar campaign, an EBX campaign, or a boot camp, irrespective of the campaign tactic, they will all get equal attribution credit. But,” she added, “If a person is added to an executive briefing, (then) a boot camp, and they got added to an EBX right after, what is the likelihood of an opportunity to close? We find patterns between different campaign tactics. That’s your key to success.”

Marketing and sales attribution evolution

As Nutanix built its attribution processes over time, it realized that change was inevitable. The company acquired other firms and also grew on its own. Growth required changes to the campaign hierarchies, and the attribution model had to be scalable. “No attribution model is perfect from the beginning,” Saumya said. There is constant tweaking, changes, and fine-tuning that needs to happen. You run your model for some time, you analyze the data, and then you need to go back and tweak it differently.”’

In the long run, the effort was worth it. Nutanix solved the old problem of the battle between marketing and sales by proving that, in reality, revenue comes from a partnership between the two teams. Saumya is pleased with the results. “We were able to define those tracks, and then we could align better and work more cohesively with our sales teams even though we don’t have one big revenue operations team. Everyone’s working in conjunction with each other—based on the attribution model.”

Recommended resources