Sales is an area where GenAI excels at augmenting human talent, in small and large companies alike.

By deploying AI capabilities throughout the sales cycle, businesses can increase sales productivity by 50%.

This increase can be achieved by increasing funnel throughput (more qualified leads), increasing conversion rates and retention rates (personalized automation), and reducing time spent on low value activities. What’s more, by using GenAI for solution development, it’s often possible to start demonstrating impact within 3 months.

Read this post, and you’ll know exactly how to get started.

Identifying the sales pain points

The first step is to ask marketing and sales teams about improvement areas, as they usually have a good sense of where the company is leaving money on the table.

Complaints that are often mentioned include:

• Lead generation activities struggle to break through in an increasingly noisy marketplace. Whether it’s for outbound marketing campaigns, lead magnets (gated content) or account-based marketing activities, an ever-growing amount of work is needed just to maintain the company’s share of voice and lead quality.

• Lead qualification is slow and not customer-friendly. Usually, lead generation activities generate large contact lists that are handed over to sales development representatives (SDRs) for manual qualification. While there has been tremendous progress in the use of lead data aggregation platforms like LinkedIn Sales Navigator and Apollo, lead qualification invariably goes through a stage of sending follow-up communications to contacts. This stage is time consuming for salespersons and viewed as unpleasant spam by recipients.

• Salespersons spend too much time inputting data into CRM tools during the sales cycle. Key account managers spend 5-10 hours/week updating platforms like Salesforce and bringing stakeholders up to speed, or hire sales assistants to do the work.

• Product onboarding is painful for customers, hurting retention. Customers must often re-input data that they have already provided. At worst, they must undertake significant manual work to migrate their data into a new system.

• Sales/product feedback loops are imprecise and opaque. Customer feedback is essential for product managers to iterate on new features, but it is conveyed sporadically by salespersons and often biased by whatever large deal they are focused on at the moment.

GenAI quick wins

Companies can address many of these pain points by implementing AI-powered sales automation systems, and they can do it quickly by using AI for platform development.

The following diagram summarizes potential quick-wins, each of which can be prototyped in just a few weeks.

Creative GenAI tools for content generation

This AI use case is well known, thanks to the tremendous adoption of ChatGPT (400 M users and counting). Often, a small number of young, passionate prompt engineering experts emerge organically in each organization.

The main implementation challenge is to ensure that the vast majority of sales and marketing employees are trained to properly use chatbots to brainstorm new content ideas, produce high-quality copy and images, convert content into multiple formats (e.g., animated banners, video explainers, Instagram and Tik Tok clips), and even monitor key opinion leaders on social media.

AI-powered data population

A specialized AI agent can perform web searches, scrape and parse web pages on the contact company’s website, and query third-party APIs (e.g., Apollo, Proxycurl, SimilarWeb). It concatenates all that information to extract and format relevant pieces of information such as company turnover, number of employees, fund raising history, web traffic and number of published job postings.

Personalized lead follow-up

In a recent post, I talked about how to use an internal knowledge base of customer success stories to generate personalized follow-up emails based on the recipient’s industry sector and company department.

This concept can be further enriched by referring to the contact’s activity on LinkedIn or Twitter/X to personalize messages, for example.

New CRM paradigm: from structured to unstructured data entry

With AI-powered CRM, salespersons can dump information into the CRM’s lap and let it organize, format and store it. Similarly, when they need information about a company or contact, they can ask a single question and get all the relevant information displayed in a neat dashboard. No more navigating drop-down menus and filling multiple form boxes.

Companies can save hours of weekly data entry tasks by improving CRM workflows. Ideally, the AI-powered workflows should be deployed “on top” of existing databases and CRMs and communicate with them through APIs, so that no data migration is needed.

Sales co-pilot

The sales co-pilot is an AI agent that continuously thinks about what’s the best next action that the salesperson can take based on the state of their pipeline. It generates a first draft for the salesperson’s review, such as:

• An email to a customer contact.

• An outreach to other potential decision makers at the contact’s company.

• A question to an internal stakeholder, or an internal approval request.

• A progress report for the salesperson’s manager.

When deploying co-pilot agents, it’s important to keep things simple at first, to avoid swamping the sales team with distracting signals. A potential approach is to focus on creating, say, only 3 relevant draft emails per day, and measure adoption.

AI-powered customer data migration agent

An AI-powered customer data migration agent aims to reduce the amount of manual work required by the customer to fully migrate to the product that you have just sold to them.

Its specific implementation depends on the context, but at a high level, the agent’s role can be broken down into two main steps:

• Planning: Based on what it knows about the target system’s data structure, and by reviewing the customer’s existing systems and data, the agent can design a migration workflow and even generate code snippets to map data from one format to another format.

• Actual migration: The agent can run migration scripts and call on very low-cost AI assistants to perform data mappings that require language interpretation. For example, mapping a “Technology & Software” industry segment to either “Technology hardware” or “Technology software” based on other company information.

AI-powered sales/product customer feedback loop

The goal of the customer feedback agent is to aggregate customer comments from all sales interactions and generate feature requests for the product manager, together with a quantitative assessment of the number of accounts with explicit or latent demand for each feature.

The agent’s workflow is fairly straightforward, closer to that of an AI assistant than an AI agent:

• Step 1: Read and summarize all customer insights from sales conversations (either verbatim or interpreted by the salesperson) and generate a first list of feature requests.

• Step 2: Based on the list of feature requests generated in Step 1, check all available sales conversations to score whether each particular account is likely to be converted or retained if this feature is added.

Takeaway messages

AI-powered sales automation can transform traditional sales processes into more efficient operations that leverage the best of what humans and AI agents have to offer. Quick-win solutions can be prototyped and implemented within weeks, as long as the company is prepared to remove them quickly if the organization’s feedback is poor. Targeted improvements, such as the ones described in this post, should be preferred over full-fledged re-engineering of sales workflows.

In future posts, we will discuss and demo most of the AI agents discussed today.