Analyst Perspectives

I have written about vendor efforts to use artificial intelligence (AI) and advanced analytics in their applications targeted at sales and revenue teams to improve focus and prioritize activities, both for pipeline management as well as individual opportunities. Since then, vendors have continued to innovate, and there have been more releases showcasing efforts to aid sales and revenue. And with this continuing innovation, we believe that by 2026, two-thirds of revenue leaders will begin considering a new generation of revenue analytics and data-driven applications designed to improve performance and productivity.

Some recent developments have focused not on deals and deals in the pipeline (more on this later) but on the origination of deals. Traditionally, when it comes to leads, the strategy has been to generate interest through website SEO for inbound and rely on sales and business development reps (SDR/BDR) for outbound against targeted accounts. This is very much driven by the idea that the more leads you have, the better the odds of having a lead that results in a qualified opportunity that can result in a sale. And although this is logically correct, it means a lot of effort and resources go into qualifying most of these leads and result in either a flat-out rejection (not real interest, no budget, wrong type of company, etc.) or diversion to a nurture program (no budget yet, months out, early-stage project). Marketing and SDR are involved in this effort, aided by rules or machine learning (ML), resulting in a score that is used to triage the deal. But even then, leads convert to opportunities that, for good reason, do not progress much beyond early-stage activity. And I haven’t even mentioned “zombie” deals that are kept as part of a rep’s pipeline just to ensure coverage ratios are met.

One way to improve this lead-to-opportunity flow is to, in a sense, reverse the process by qualifying leads “in” by identifying accounts with a higher likelihood of having an opportunity through better product/market fit. This is the idea behind account-based marketing which is now evolving into a more granular approach as technology is enabling this practice to be taken a step further. Rather than selecting accounts based on a broad product fit predicated on similar successes in the past, newer applications are utilizing more advanced analytics to examine organization-level public financial data and news stories and analyze ratios against benchmark data to identify where there is potential for process and technology improvements. This will self-select accounts based on a vendor’s capabilities, depending on the product and services that they offer. Going further, having identified potential prospects, auto generation of sales materials will focus on how the vendor, through their product and service offering, will deliver incremental value for the prospect. Though this is early days for this approach, it shows additional promise as to how utilizing data and advanced analytic techniques can help organizations improve the conversion rate of lead to opportunity by self-selecting prospects with a preset value offering rather than utilizing resources to qualify leads and opportunities.

Another area where recent technology development is being utilized to assist in sales and revenue is enhancing support for remote selling. While there are many productivity and cost-saving reasons to continue with the shift to remote selling, this is often at the expense of losing some of the more human signals that salespeople rely on to gauge the reception of messages, presentations, demos, and quotes. Advanced analysis of conversation has been around for several years, using AI/ML to analyze responses for positive and negative sentiment as well as monitoring share of conversation. But this approach, although of undoubted value in analyzing whether messages resonate or not, cannot decipher the additional signals humans display through facial expressions and body language. While virtual meeting applications can provide the logistics of an in-person meeting, they cannot easily make visible the facial expressions and body language that would otherwise be readily available. Levels of personal engagement, distracted behavior, negative and positive expressions and body language such as leaning in are much harder to discern via a small visual window representing multiple attendees. Vendors have recently announced new releases that showcase the capability to complement conversation analytics with facial expressions and body language analysis, synched to the conversation and presentation from the vendor.

As with any new feature utilizing AI/ML, the key to delivering value is less about the technology or the techniques employed and more about how it can be used in existing processes. If the processes need to be adjusted, those charged with deployment within an organization need to understand that it is not just about the what and the how, but more importantly the why. As an example, vendors talk about feedback in real time, with both conversational as well as the newer facial expression analysis, but the reality is that very few salespeople, or anyone else presenting, can both focus on a meeting and react to indicators in real time without appearing disjointed or unfocused. A better use of this technology is in analyzing after a call as part of either an individual coaching session or as part of general messaging, presentation, demo, or quote reception analysis. The aim should be to utilize this feedback mechanism to improve overall interactions and fine tune messaging and value propositions. And a final word of caution: Many of these applications that utilize approaches such as ML are heavily dependent on having quality datasets that capture what “good” looks like versus “bad” to train models that are reflective of recognizing what behavior or actions lead to a successful outcome. It is always worth validating with any vendor as to whether your own data provides this information in terms both of quality as well as quantity in terms of volume of data and in terms of length of history.

For those open to using data and AI/ML to improve marketing and sales effectiveness, these initiatives are worth looking at, with the caveat that there is no magic bullet. The market is littered with great ideas that foundered on deployment because of a lack of internal focus on the why, with too much focus on the what, especially for technology intended for sales and revenue teams.

Regards,

Stephen Hurrell