The Past and Present for AI in B2B: 4 Ways AI Can Help Companies

By Kerry Cunningham, Senior Principal, 6sense

When I assert that the first artificial intelligence devices were clay, not silicon-based, you would be justified in asking yourself why you haven’t heard of these clay-based computers. The answer is that the first artificial intelligence devices were not computers, but tablets made out of clay, upon which people recorded transactions, directions, genealogies and a wide variety of other important but hard-to-remember and, let’s face it, mundane things, beginning about 5,000 years ago.

How do clay tablets qualify as artificial intelligence? The answer is that they, like any form of writing and most computers, serve as external storage devices for information, which can be retrieved and used to settle accounts, duplicate receipts, repeat complex journeys, and aid in virtually any task that would otherwise overwhelm the typical human mind. Clay tablets were certainly more portable than the first mainframe computers, although probably substantially less fun to carry around than an iPhone mini.

And if keeping tabs on who owes whom for what doesn’t sound like the coolest first use of artificial intelligence, consider what life would have been like before this first form of writing was invented. Absent the ability to externalize memory, much of the vital information compiled by individuals, tribes, bands and even larger communities would have frequently been lost or distorted as the people carrying the information forgot, aged, died, or just misremembered. And that’s not to mention how much squabbling must have been going on all the time over who borrowed how many of this for that, or who traded how many chickens for how many goats last year.        

Moving information out of our heads and onto a more permanent and dependable media than our brains was key to accelerating cycles of innovation that have led us to where we are today. Externalizing information storage continues to be one of the key benefits of artificial intelligence today.

The Problems AI Helps With in B2B

Below, I will describe four capabilities of modern AI that are particularly useful for B2B organizations, and which all B2B organizations should be taking advantage of now. But before I get to that, let’s pause for a moment and consider the significant problem nearly every B2B organization must address. Simply put, it is to more accurately identify and prioritize potential buyers. Today, most marketing organizations are still focused on producing leads – individuals who have demonstrated some interest in the company’s solution. What makes this an urgent and important problem to fix is that B2B lead conversion rates are nearly always below 5% from initial lead to closed won business. For B2B organizations that sell departmental or enterprise scale solutions, that metric is typically less than 1%. So, clearly, marketing organizations have not nailed the identification of actual buyers yet.

That metric also provides context for sales productivity. During my time as an analyst, we typically found that marketing supplied fewer than 20% of the selling opportunities sales eventually worked, leaving sales to do their own prospecting for the remainder. Virtually no one actually knows how productive sales prospecting is, because even today very few companies have anything like a complete record of sales rep prospecting activity. Without the complete record, nearly everyone relies on intuition and experience. Good enough, you might think, but if your competitors actually work from real data to know what their sales productivity really is, will your intuition and experience fare well? As we will see, AI comes to the rescue both in helping generate more sellable prospects, and in collecting sales productivity data.

The Four Ways AI Helps in B2B

Critical AI Use Case: Signal Stitching

Perhaps the most fundamental way AI helps is by capturing and stitching together more data – more signals from buyers, which can be used later to more accurately determine which are in good near-term prospects and which are not. Some of this data sits inside internal systems such as marketing automation, CRM, billing systems and the like, but much of it is distributed across the internet. There are three broad categories of buyer signal that must be collected and stitched together.

1. Fit Signals. For most organizations, the types of accounts that are a good fit for their solution are determined through the use of standard firmographic and demographic variables. For company fit, these variables typically include industry, size, and location variables among others. For person fit, these are typically role, level, department type

2. Intent data. As buyers research solutions on the internet, they interact with a wide variety of web sites, forums, social media sites and the like. Gartner estimates that only about 17% of buyer journey are spent on vendor websites, so that leaves a lot of traffic to be found and understood from outside any given company’s digital domain.

Even on each company’s digital properties, there is a substantial volume of buyer signal that is largely ignored. Most organizations experience conversion rates of unique visitor to form fill lead of between 1% and 5%, meaning between 99% and 95% of all visitors remain anonymous. Organizations must do everything within their legal power to identify the source of that information and associate it to both the third party intent signals just described, as well as their own leads

3. Current State Signals. Forrester describes current state signals as signals that depict the current technographic, economic, regulatory and financial condition of an organization. This category of signal is critical, because a company’s current state is typically what causes a company to go into the market to look for business solutions. For example, a company that has just received a substantial round of funding has likely just entered a state in which it will need additional technologies to help in the recruitment, hiring and training of the new employees. Similarly, companies in highly regulated markets may face a need to acquire new technologies when relevant regulations are changed or added in the markets in which they operate.

4. Leads data. The leads companies typically receive and house in marketing automation systems are another source of signal that must be stitched together with the other signals mentioned above. During my time as an analyst, I had a first-hand view of hundreds of B2B organization’s leads data. The insight from that experience is that most organizations receive multiple leads from organizations that eventually become buyers. However, in most cases, only the first lead from any given account would be associated with the opportunity. Without stitching those leads together with the other signals mentioned, organizations are effectively ignoring critical buying signals.

The B2B ecosystem is creating enormous volumes of data, but without being able to stitch it all together - tying the current state and intent data together with firmographic data to create a coherent picture of a potential buyer, its needs and intentions, all that data is not particularly useful. The job of figuring out which signals should be stitched together to form a single buying team signal is one of the principal use cases for AI in B2B.

Critical AI Use Case: Pattern Matching

Stitching all that data together is foundational, but not, in and of itself productive. Once the stitching is in place, real productivity enhancements for marketing and sales come from identifying patterns in the data that identify     and even predict which prospects are buyers (or will be) and which are not. It is not as simple as noticing when relevant organizations are demonstrating interest in a category of solution. In B2B, where people in organizations have specialized roles in HR, IT, manufacturing and the like, virtually every organization will be exhibiting signals of interest in relevant solution categories all the time. For example, a company with a 100 person HR department can be expected to have at least a few members of that department visiting HR software and industry publications numerous times every day. It takes the pattern identification power of AI to notice when that pattern of behavior looks different from normal – and when it is accompanied by the right business conditions (e.g., company is going through a growth spurt, company has a current HR software suite that is not appropriate to the size company it is becoming).

About the author: 

Kerry Cunningham, Senior Principal at 6sense, is driving the buying groups revolution in B2B, helping organizations transform from ineffective lead-based practices to modern, buying team and opportunity-centric processes to unlock next-level performance. As a former SiriusDecisions and Forrester analyst, he has been a leader in the design and implementation of demand marketing processes, technologies and teams for a wide array of B2B products, solutions and services. Kerry brings to his work with B2B organizations a unique combination of academic and real-world expertise in marketing, organizational design and management.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

Other Topics

Technology