B2B Marketers Adopting AI To Streamline and Scale Personalization, Messaging And Consistent Customer Experience

Artificial Intelligence (AI) and machine learning are being baked or integrated into every aspect of B2B technology. Early adopters are seeing success leveraging AI to enhance and scale personalization efforts, offer content recommendations via owned channels and streamline the sales cycle.

Research from Demand Gen Report and Demandbase supports this claim, showing that nearly 80% of B2B marketing and sales professionals feel a significant amount (more than 20%) of their current marketing and sales applications will be AI-powered by 2020. Additionally, 70% feel AI-powered applications will help improve and accelerate the buyer’s journey by recommending next best actions.

“In addition to having AI help with targeting, AI is helping drive better performance when engaging target accounts in areas like ad buying [analyzing and optimizing layout, creative, copy etc.], content customization and web personalization [industry/account/role-specific content, next logical asset, etc.],” said Matt Senatore, Service Director for Account-Based Marketing at SiriusDecisions, in an interview with Demand Gen Report. “It also helps identify which accounts might be at risk of defection.”

Experts noted that the marketing department, in particular, is where AI has the biggest potential to impact B2B businesses. The study noted the following roles have the greatest potential to benefit from AI-powered applications:

  • Demand generation (61%);
  • Digital marketing (56%);
  • Marketing Ops (45%); and
  • Customer experience (40%).

Qubole is a company that has seen its trial signups increase by 292% from leveraging AI to personalize website messaging and provide relevant content recommendations. With these types of AI use cases coming to light more frequently, vendors in the space expect to see AI continue to adapt to answer specific pain points B2B organizations have regarding content production. While the technology isn’t there yet, emerging solutions with content recommendation algorithms are expected to suggest the types of content that should be produced to meet the needs of their target audience.

“With AI, you can get specific on what the account wants in their content,” said James Regan, CMO and Co-founder of MRP, in an interview with Demand Gen Report. “We think the next place is AI-enabled content providers. That’ll be the next area where organizations such as MRP will bake in AI because that problem is real. You can only create so many content matrices; there will always be something needed — it’s beyond what a normal size company or agency can cost-effectively create.”

Early success from AI adopters show that B2B organizations will begin to adopt this technology rapidly to keep up with buyer expectations and their competitors.

“In 10 years, there will be an AI-driven marketing stack that will cover the basic common processes, with abilities to extend and augment those models,” said Aman Naimat, SVP of Technology of Demandbase. “You can’t have personalized processes without machine learning in the data.”

Qubole Finds AI Success With Content Recommendations, Account Personalization

The most discussed use case for AI today is the ability to position relevant content on a company’s website for prospective customers and accounts based on past engagement.

For example, Big Data service provider Qubole integrated AI applications from Demandbase into its website to offer personalized content recommendations and other messaging to IT engineers in a notoriously noisy market. The company needed to target a narrow set of personas with personalized messaging to alleviate the often long, complex buying cycles experienced in the Big Data industry.

“[Qubole’s] challenge was to identify companies at different levels of maturity and then develop personalized messaging across multiple points in the buyer’s journey to reach their target accounts,” said Naimat. “Qubole was able to implement Demandbase’s AI-enabled account identification and targeting to optimize and scale their customer experience to accelerate the buying process.”

In addition to using the platform for account identification and targeting, Qubole also uses Demandbase’s account intelligence and website personalization offerings. This positions the company to make educated content recommendations based on the prospect’s past engagement behavior.

Ultimately, Qubole saw the exit rate on its website decrease by 39%, while the average time spent on site increased by 467%. The company also saw the number of signups for a trial increase by 292%.

Predictive Capabilities Evolve Into AI Algorithms

Experts noted that AI is already having a significant impact on marketing and sales technology (and marketers/sales leaders) today — except some might not associate AI as the driving force behind this. A notable buzzword in the B2B marketplace over the past five years was predictive analytics, which involves leveraging technology to predict prospects and accounts with the highest propensity to buy. This is considered a form of AI based on past purchase and engagement data from closed business.

“Many companies are using predictive analytics solutions for a variety of marketing and sales use cases today,” said Senatore. “Predictive analytics solutions are building algorithms and continuously refining them based on more and newer information.”

Senatore added that this form of AI uses machine learning to improve the predictions it makes around:

  • New accounts to target;
  • Existing accounts to prioritize; and
  • The next best offer or right solution to suggest into a new buying center.

Matt Amundson, VP of Marketing and Sales Enablement at EverString, stated that predictive was the first tool in B2B sales and marketing to leverage the term “AI.”

“Predictive was one of the first categories in B2B sales and marketing to start leaning on the term AI,” Amundson said. “Predictive introduced the idea that a piece of software could look at successful outcomes and make recommendations to optimize a business around creating more successful outcomes.”

Amundson pointed to SalesLoft, a sales engagement software provider, which uses EverString “to model their customers to predict which of them have the highest potential to add licenses.” He added that SalesLoft “leverages that data specifically for post-sale activities and engagement.”

Closing Data Loops To Connect The Dots With AI

An ongoing challenge discussed among B2B practitioners regarding AI adoption is the hesitancy to rely on decisions made by algorithms fueled by inconsistent, inaccurate or old data. Research from the Demandbase and Demand Gen Reportstudy showed that close to half (46%) of respondents said that they are holding off on adopting AI because they “had a lack of trust in decisions being made without human oversight.” In addition, 42% said they “feared a general lack of control” in the decision-making processes of AI.

“AI isn’t calling the shots yet,” said Allison Snow, Senior Analyst for B2B Marketing at Forrester Research, in an interview with Demand Gen Report. “There has to be a lot of human oversight. You want to do a logic check to review the data and review the output. It’s worth digging deeper to better understand what data it deemed most relevant.”

Snow suggested that, to properly fuel an AI algorithm with the insight it needs, B2B marketers should:

  1. Review the data that will likely train this model; and
  2. Ensure the model has access to that data.

“If we continue with an example of using an AI-powered [algorithm] to predict churn, a B2B marketer should know how accurate the data records are that indicate churn,” said Snow. “Has the organization been disciplined in documenting the reason for churn, as reported by customers? Is there a drop-down list in a structured field, in which a salesperson or customer service rep is expected to provide a reason for non-renewal? If so, have they been using that field consistently? Or have they carelessly just selected “non-renewal” from a drop-down and moved on?”

Snow added that the data points marketers plan to use are worth an “audit” of some kind. Then, the algorithm must have access to that insight. “Again, is it in a structured field or non-structured? If it is non-structured, does the tool have the specific AI tech to interpret the data, like text analytics that can learn from unstructured data?”

Ultimately, B2B marketers need to know the formula for model-making with machine learning is:

Data + algorithm = AI model

“Algorithms are inherently objective,” Snow concluded. “Therefore, machine learning models are only as good as the data you use to train them.”

This article was originally posted Demand Gen Report


Mark Halstead

You might also be interested in