How to Fuel AI-Powered Marketing
While artificial intelligence’s capabilities have expanded greatly in only a few years, to get the most out of the technology, marketers need to cast a wide behavioral data net, suggests Steven Casey, vice president and research director at Forrester Research.
AI is becoming increasingly important in sales and marketing because it can orchestrate complex omnichannel engagement across the buying cycle, in many of the leading customer data, account-based marketing (ABM), and marketing automation platforms, Casey writes in a new Forrester report on the relationship between behavioral data and AI.
Companies should ramp up their deployment of new AI-powered functionality, which can now drive programmatic advertising, predictive analytics, measurement, virtual assistants, chatbots, marketing automation, customer data platforms, ABM platforms, site personalization, and sales enablement personalization.
“AI-powered systems require accurate, updated, and complete data, preferably as detailed and diverse as possible, from a wide variety of new (or underutilized) first- and third-party sources,” Casey says. “This will be a significant challenge for many organizations.”
Forrester found that nearly one-third (32 percent) of global marketing decision makers identified managing data quality as one of their organization’s biggest challenges. While some customer information is necessary to establish identity, this information, on its own, is not enough to enable deep personalization, according to Forrester. This is where behavioral data comes in. Marketers should concentrate on how customers, prospects, and influencers engage with websites, email, advertising, peers, and sellers throughout the purchase life cycle, Casey advises.
The statements that customers make regarding their needs, plans, preferences, and sentiments are among the most valuable, but even inferred data can provide valuable insight into customers’ intent to purchase. That data can drive AI with additional input to next best recommendations or triggered activation.
Before going all out with new AI functionality, Casey recommends a few initial steps for marketers:
- Take stock of current and desired behavioral sources. Once current data sources are identified, look for gaps or unmet needs, then create a comprehensive road map to achieve the company’s business goals and objectives.
- Use a pilot before a broad rollout. It’s important to build the business case via a small trial so that you know the solution provides the desired results before going with a broad rollout.
- Name the pilot project. Internal stakeholders in sales and channel production don’t care about AI or data management, while management cares only about results. So the best way to get on the same page is to refer to the pilot project by a fancy name, like Project Zeus.
Forrester also identifies the following three key elements for developing a holistic and comprehensive behavioral data strategy:
- Pay attention to what prospects do rather than what they say. Rarely do prospects ask directly to be contacted by sales. So Casey recommends marketers focus their efforts on the actions of target accounts and contacts, from the first anonymous interaction to post-sale product usage, expansion, and renewal.
- Cast a wide net to collect as many behavioral signals as possible. The report suggests that marketers collect behavioral data from as many as five sources. An increasing array of third-party providers can augment companies’ own data sources; many of these data solutions also can provide a more granular set of data signals via additional services, the report points out.
- Create a single source of truth. Casey recommends that marketers consider investing in an internally developed data management solution or a packaged application to cleanse, deduplicate, and normalize their customer data and to build unified profiles that incorporate all the assigned, observed, and inferred data available on target and existing accounts and contacts.