Views from the Marketplace are paid for by advertisers and select partners of MIT Technology Review.
Professional services firms see huge potential in machine learning
Business-to-consumer (B2C) businesses have made it a priority to incorporate machine learning into customer-facing functions, integrating it into sales and marketing.
For business-to-business (B2B) companies, however, translating data into actionable marketing strategies can be a more difficult proposition. Selling to organizations invariably requires embarking on a much longer and more complex journey, culminating in an order of much higher value than in the consumer realm. With hundreds of thousands, if not millions, of dollars at stake, a misguided marketing investment could lead to financial losses.
“The availability of data and the importance of having the focus on the full customer journey is coming a little later to the B2B world,” says Laura Beaudin, a partner at Bain & Co. “A lot of expectations in terms of customers manifested themselves in the consumer world before they brought those expectations to their business-purchasing world.”
Even so, marketers in the professional services sector are determined to master machine learning (ML) and other tools to drive campaigns using data and insights. In a recent survey of 1,419 marketing executives, conducted by MIT Technology Review Insights in association with Google, the professional services sector—which includes systems integrators, third-party consultants, and technology advisors—ranked among the top handful of industries that are adopting and applying ML and data analytics.
In the survey, respondents in the professional services sector are 16 percent more likely to believe that predicting customer intent will drive greater marketing results than marketers in other industries, including retail, financial services, and travel—where available data is hardly in short supply. While professional services marketers may face more challenges in that regard—B2B data is generally not as plentiful or as accessible as B2C data—they haven’t allowed such early-onset obstacles to diminish their commitment. In applying technology to engaging and anticipating customers in wholly different ways, they are rethinking their strategic priorities, retooling their capabilities, and re-emerging as savvy competitors.
Breaking the marketing mold with machine learning
Finding a differentiator in data
A majority of B2B marketers are aware of the potential impact that data-driven marketing could have on improving their competitive position. In the survey, 58 percent of marketers from the professional services sector agree that how companies apply their data will play a key role in their ability to thrive.
Guided by advanced ML algorithms, B2B businesses have the opportunity to use the information they collect and analyze to gain insight into what drives customers’ behavior. As ML technology keeps refining its analysis—“learning” more about customers as it familiarizes itself with the data—it equips marketers to produce increasingly effective marketing content in real time.
Marketers must begin the process by creating a data backbone to support such innovation. That means standardizing and unifying all the data they’ve been storing, making it easier to access and harness. “If marketers knew 10 years ago that they were going to use all this data they were accidentally collecting, they would have kept it in one place,” says Jay Bowden, Google’s industry director for the technology B2B vertical. The task becomes even more challenging for companies that have grown inorganically—acquiring, along with other businesses, multiple customer relationship management systems (CRMs) and other silos. “It requires finding a common place to store the data from the silos before finding a way to make some decisions from it,” says Bowden. “Then you can let machine learning look at it and find some commonalities.”
The adoption of cloud technology has created a data repository on the needed scale. It also provides a platform for ML algorithms to process that information without compromising security.
Of course, data doesn’t automatically flock to a central location; its human guardians must be persuaded to let it go. “Even once a marketing department has a handle around their data, they need to have an internal road-show process to get other departments to hand over data,” says Beaudin. Cross-functional employee teams, she adds, need to become “better consumers of data”—applying insights derived from data to cultivate higher-quality leads and boosting conversion rates.
Initial lessons in machine learning
No matter where they stand in integrating the technology, professional services marketers know how much the marketing function can benefit from ML. In the survey, nearly two-thirds of marketers (64 percent) from the professional services industry believe using ML will allow their companies to gain a competitive advantage.
The technology can improve the marketing function’s performance by automating processes. The place to begin, says Bowden, is to “automate some marketing practices first, such as bidding” as well as using ML to test creative efforts, writing ads and applying technology to optimize them.
By using ML to make internally visible gains, marketing can also help secure a profile of sufficient altitude to help procure additional resources from management. “Historically, the leads that the marketing team has delivered haven’t been the highest-quality leads,” says Bowden—meaning they haven’t been leads that could be quickly converted to sales. But as lead quality improves, the sales function will become more aligned with digital marketing efforts. Bowden recalls one software-as-a-service client that wanted a more effective approach to predicting which leads were more likely to convert from a “free trial” to paying customers. By boosting its marketing spending from $1 per click to $1.25, the company was able to corral “a whole new group of customers.” The difference? The ability to “invest in predicting the lifetime value of our clients by predictive modeling,” Bowden says.
Guided by expectations
Among survey respondents, marketers in professional services are 17 percent more likely than other respondents to believe it is necessary to predict intent in order to win loyal, high-value customers. The capability to accurately predict intent and lifetime value ultimately ripples through all marketing initiatives, such as offers on pricing and promotions on product or service usability.
So far, says Sarah Travis, industry director for business and industrial markets at Google, "a handful of organizations in the B2B space are trying to predict what one year of orders are likely to be based off of somebody's first purchases. This will allow them to better reach and engage customers that are likely to be high-value."
Clearly, any reluctance on the B2B sector’s part to embrace the technology doesn’t reflect a fundamental doubt about whether ML could help those companies become more astute, and efficient, marketers. More likely, the sector’s marketers may be concerned about the maturity and complexity of the technology and whether they have access to the skills necessary to maximize its use. “The more complex the conversion process, the more difficult it is to understand through human touch what the intent of a customer looks like,” says Travis, who partners with several B2B companies. “The B2B world has a lot more to gain from using machine learning technology than any other industry.”
As the survey reveals, B2B marketers harbor abundant enthusiasm for the transformative technology—not only for what ML is capable of doing, but also for the ways it will enable them to improve their own performance. “Machine learning opens up a space where they can think about their marketing function and about reaching their customers in ways they never thought about before,” says Travis. “And that’s just a start.”
Learn more about machine learning’s impact on marketing.