Artificial intelligence (AI) for sales and marketing is shifting steadily from hype to hope to promise and now implementation. But it’s a highly uneven shift – differing sharply by company and by type of AI.
Victor Antonio, Host of the Sales Influence Podcast and author of Mastering the Upsell, estimates that vendors of sales AI have grown from 300 four years ago to perhaps 5,000 now. Investors are pouring money into these companies, and market valuations of sales AI companies have risen sharply.
The pandemic has accelerated the use of online tools to sell, and this has only increased the potential of AI to help both sales and marketing. “Now everyone is Zoomable,” Antonio argues.
But companies still need to be careful. “There’s lots of interest and investment, but also a lot of hype and unachievable promises,” argues Steven Wright, chief analyst at the consultancy Vendor Neutral. Wright sees the most AI progress in natural language processing (NLP) to extract intelligence from conversations and chat bots.
Wright says some other AI tools are challenged by insufficient data to develop their sophisticated machine learning (ML) algorithms.
AI seems best suited to marketing – with its one-to-many reach, massive data collection, and substantial automation already. Marketing’s crucial job is to maximize sales by getting plenty of plenty of great leads for reps, highly relevant information on these leads, and persuasive material to support reps. Reps must exploit these assets quickly and effectively.
How can AI help align the two crucial revenue departments?
Scratchpad founder Pouyan Salehi stresses that sales and marketing must align objectives before considering alignment tools. His Marketing VP Nate Odell – veteran of several startup marketing campaigns – stresses that personal relationships between reps and marketers must also be built.
Yet AI has already been transforming marketing. Forrester notes machine learning, which improves its algorithms automatically as it gains more data, has optimized media strategies, detected ad frauds, and replaced manual efforts with rules-based automation.
Marketers’ social listening tools use natural language processing and machine learning to understand buyers. Then natural language generation (NLG) can automatically generate or customize content for either marketers or reps to use with leads and prospects.
Marketing chatbots can ask and answer questions, gathering information. Chat bots can also understand the implications for sales of the questions asked by leads. And they can automate repetitive tasks like setting up and coordinating meetings, responding to leads, and sending emails.
All this makes marketers more efficient. It should also mean that marketing passes more leads that are better qualified and more valuable to sales reps.
So AI is not only aligning sales and marketing; in some cases, it is fusing the two departments. Antonio sees several companies cutting investment in field reps and spending the money on inside reps who will work in marketing departments, so they can pounce quickly on leads generated by AI-assisted marketing Websites. “Before they pass the lead to sales, their own reps can call and say, I saw you looked at our video.”
AI’s power to track and analyze individual behavior online is exploding. Modern Internet tools can detect the scrolling speed of a visitor checking out your Website. They can record not only which items the visitor downloads or buttons he or she clicks on, but where and how long the lead’s mouse hovers. “All these are signals on what the visitor is thinking,” Antonio explains. Then AI can interpret these signals to tell marketers and reps both the likelihood of a sale and the best approach to achieving it.
Antonio argues all this data and its valuable fruits are not limited to a seller’s own Website. Modern Internet tools can track an individual buyer across the Internet – noticing which other sites are visited and for how long. If these are Websites of your competitors or partners, this too provides valuable information that AI can use to assess the possible pipeline value of the Internet traveler.
Another rich data source is social media, which can be combined with Website and other data to profile leads. Here too, AI can use NLP to understand and interpret the text of a lead’s comments as they might relate to interest in your company’s products. “AI can look at 20,000 signals before the sales call,” Antonio emphasizes.
So AI can help marketing both gather leads and set the priorities for sales action. Both the quantity and quality of leads should be improved, and reps should get better ideas on how to approach each lead.
“AI has the potential to transform sales and marketing by identifying the priority market segments and target accounts to pursue,” argues Greg Hessong, an executive advisor in Gartner’s sales leader practice. AI tools can thus make or aid decisions on where sales leaders should invest time and resources.
Because AI can unify sales and marketing teams, Hessong sees it as improving sales and marketing alignment and gaining insights for better decisions. He calls this “guided decision-making”: identifying the best leads or prospects to pursue by analyzing likelihood to buy, probable revenue or profit, higher lifetime value, and the likelihood that a new account will “stick” – be retained long after the initial sale.
For marketing, AI’s lead and segment scoring can help prioritize spending on advertising and other digital marketing campaigns.
Friction between departments can be dicey, but good data from AI can smooth out ruffled feathers. “AI will help align sales and marketing if all agree they can trust the data,” Wright agrees. “It can cut out a lot of finger pointing.”
AI should also help improve sales enablement tools like white papers and presentations, and customize these materials for individual prospects. Antonio argues AI can even generate sales content automatically online.
Antonio believes that AI will eventually strengthen the role of marketing in not only finding leads but pushing them further down the sales funnel toward closed deals. In many cases, reps might come in only toward the end of the sales process.
And if a company’s products eventually turn into commodity-like offers, then sales reps could become the chief differentiators – showing empathy and other human-only qualities to a prospect who is almost ready to buy. In this scenario, fewer reps might be necessary, but their efforts would be critical and high value.
Wright too believes AI can observe prospect behavior better before reps engage them, using chatbots to elicit more information. This should both produce better leads and move the lead further along the buying process. “It’s all about scoring,” he says. These bots should be able to advise how far into the sales process a lead can progress before it needs a rep.
How far are we on this journey to AI-aligned sales and marketing? Antonio estimates only about a quarter of B2B sales organizations are using AI for any purpose at present.
For beginners, Antonio recommends starting with a very specific sales problem to be solved and then focusing on deploying only a couple AI tools at first.
Always remember the data support upon which effective AI relies. Unstructured data (such as conversations and images) must be converted to structured data that AI can manipulate, like math equations manipulate numbers. And structured data – real numbers and clear terms – must be collected and aggregated in a single data platform.
Some, but not all, of this data may already be in well-kept CRM systems. The rest may have to be pulled together from different databases.
Wright warns that most B2B sales organizations simply do not generate massive sales data as rapidly as B2C companies like Amazon – so it may take several sales cycles to accumulate data sufficient to support some AI tools. He thinks medium-sized companies may already have more data than managers realize. “For large companies, data is in so many systems it is more difficult to pull it together.”
Of course, one partial solution might be to find a way to aggregate and anonymize all the data sitting in Salesforce, which would be an incredibly rich database for guiding B2B selling. But Wright says nobody is attempting that yet.
Hessong argues sales leaders’ disappointment with AI usually results from one or several causes: poor use case selection; trying to solve every problem; poor or missing historical data; inadequate training of marketers and reps in using AI; or unrealistic expectations of what AI can do.
“You’re not going to win 100% of the time,” Hessong stresses. “And you have to help reps and marketers understand how to embed AI in their daily workflow; otherwise, they won’t use it.”
That sounds a lot like the early (and sometimes chronic) problems with CRM systems. Of course, no technology can win over every rep on every team. Some markets will adopt AI faster than others, especially when there’s solid evidence that sales will improve and marketing will help push sales forward.