n accurate sales forecast requires a crystal clean sales pipeline. A crystal clean sales pipeline requires a deep history on most buyer/seller interactions for each opportunity. By having a deep history of these interactions you can apply AI to identify buyer patterns and better guide the sales process. Unfortunately most sales organizations do not have enough information about each opportunity to apply AI to assist in the sales process.
Gartner calls this out as one of the impacts in their research “Optimize Sales Execution With Artificial Intelligence for Guided Selling, 2019” by stating:
“AI functionality requires that selling organizations improve the quality of their sales data before introducing new forms for guided selling to their companies”
Enhanced data collection is foundational for both the sales process in general and when applying AI to the sales process.
The Current CRM/Sales Rep Environment
The current environment between the sales rep and there CRM solution reminds me of the quote: “the beatings will continue until morale improves”. The current environment is doomed to failure. It is manually intensive. Updating the CRM system by the sales reps is done at the expense time they can be in front of the customer selling. In the current environment, there is nothing in it for the sales reps to keep their CRM system up to date, no incentive. They update the CRM system so they can be micromanaged by their managers. Ultimately, even if they do religiously update their CRM system, their updates are biased by a “happy ears” vs “sandbagger” disposition.
The solution is to have a system to collect all the buyer/seller interactions as a by-product of what the sales rep does. It needs to work where the sales rep works, in their email, calendar and on their phone.
The collection of email content and calendar information is accomplished via an addin in the email/calendar agent. The addin:
Automatically collects emails relating to opportunities and aligns them with the opportunity.
Automatically collects meeting from the sales rep calendar and align them with the opportunity.
Allows the sales rep to do most of their work through the addin.
Provide a mobile interface so sales reps can work anywhere, managing their alerts, dictating meeting updates and status to update the opportunity.
Three AI Use Cases
Although the data collection of email and calendar seem like a mechanical data integration problem, AI can help align and interpret the data.
The first use case of AI for the Automated CRM Data Collection can be thought of as an extension of what was mentioned in the previous blog, “algorithms to match multiple email domains and sources of information to specific leads and matching leads to companies and opportunities”. This is an area where typical CRM applications falls apart. You may have, ten, twenty or thirty leads in your CRM application from a company as the result of downloading material and registering on your website. However once an opportunity is created and a lead is converted to a contact, you lose all the interactions. The process for converting all leads to contacts is very mechanical, time consuming and detaches all the lead activity when creating a contact. This is where AI is applied to understand all of a companies email domains and loosely connect all the leads to the opportunity.
The above screen highlights an add-in for GMail that flags all emails from prospects and provides an dialog box (circled in red) allowing the sales rep to take action on the account from within email.
The second use case of AI can be applied once the first use case completed. The second use case uses patterns in history of closed-won opportunities to define the roles of a typical buying team and then applying those roles against the current opportunity leads and contacts to understand who should be pulled into meetings and be part of the team. This helps streamline the sales process by getting everybody in the room to make a decision as soon as possible.
The third use case of AI is designed to create an unbiased interpretation of the emails and calendar items by understanding the positive or negative nature of the interaction. It requires a lexicon of positive and negative sales terms used in conjunction with Natural Language Processing[NLP] to rank the sentiment of an interaction. Further AI is used to refine the lexicon by allowing sales reps to rate the emails and calendar items with a simple measure of thumbs up or thumbs down and interpret their input to update the lexicon. Further refinement could also be applied by looking at the sales rep rank (another topic for AI) and ranking the ratings higher from more experienced sales reps.