How to use AI in sales: 10 practical applications
Ten uses of AI that already work in B2B sales teams, from summarizing conversations to answering RFPs, and what each one needs in data and human review.
Short answer
Use AI where it takes repetitive work off reps, and keep a person reviewing anything that reaches a customer. The most mature uses are summarizing conversations, preparing meetings, drafting emails, prioritizing opportunities and researching accounts. All of them depend on data quality, and only 35% of sales professionals fully trust their company's data 1.
Key takeaways
- Adoption is already broad. In Salesforce's 2024 State of Sales report, 81% of sales teams were experimenting with AI or had fully implemented it 1.
- The uses that pay back first tend to free up rep time, like summarizing, preparing and drafting, or help decide the next step.
- The core risk is a wrong answer delivered with confidence, which NIST calls confabulation 3. Human review and traceable sources reduce that risk.
Where AI already makes a difference in sales
AI reached sales teams faster than most sales tools. In its 2024 State of Sales report, based on 5,500 professionals, Salesforce found that 81% of teams are experimenting with AI or have fully implemented it, and that 83% of teams with AI grew revenue in the past year, compared with 66% of teams without it 1. That comparison shows an association, and the company behind the survey sells the tool, so read the figure as a sign of direction.
In McKinsey's 2025 global survey, 88% of respondents said their organization regularly uses AI in at least one business function 2. B2B sales is earlier on the curve. In a 2024 survey of 3,942 B2B decision makers, McKinsey found 19% already implementing generative AI use cases for buying and selling, and another 23% in the process of doing so 4.
10 practical applications
1. Summarize conversations and records
CRMs like Salesforce already generate one-click summaries of accounts, opportunities, leads and contacts 9. The gain is reading time. The care point is checking names, amounts and dates before anything gets logged.
2. Prepare for meetings
AI pulls the account history, recent interactions, news and open commitments into a short briefing. It works best when the history has been logged, because AI can't know what nobody wrote down.
3. Draft emails and follow-up messages
A draft built from the context of the conversation removes the blank page. Whoever sends it checks the tone and the facts.
4. Prioritize opportunities
Scoring models estimate win probability from pipeline history. They help order the day, and the explanation of why an opportunity moved up matters as much as the score itself.
5. Suggest the next step
In a 2025 survey of 227 sales leaders, Gartner found that organizations giving reps AI-generated next best actions are 2.6 times more likely to achieve commercial growth 5. The press release doesn't detail the calculation, so treat the number as an indication.
6. Research accounts and markets
Summarizing an organization's website, publications and news before a first contact takes seconds. Gartner predicts that by 2027, 95% of seller research workflows will begin with AI 5.
7. Respond to RFPs and public tenders
In public-sector sales, much of the effort goes into assembling consistent technical, commercial and legal responses. McKinsey lists RFP responses among the uses of generative AI in B2B sales 4. AI builds the first draft from approved materials, and the team stays accountable for what gets submitted.
8. Track outside signals
Regulations, official publications and market moves can be read and linked to the right accounts. A news item that affects a customer becomes a concrete reason for a conversation.
9. Support team coaching
Analyzing recorded conversations reveals patterns, like frequent objections and the questions that move a deal forward. It requires recording that is transparent to everyone on the call.
10. Keep relationships alive outside the pipeline
Partners, former customers and contacts with no open deal tend to drop out of the routine. AI can flag who has gone quiet for too long and suggest a reason to reconnect.
How to start without wasting the investment
- Pick a measured problem, like time spent logging, slow follow-up or meetings without preparation.
- Clean up the data the AI will use. In Salesforce's survey, only 35% of sales professionals said they fully trust the accuracy of their company's data 1.
- Start with tasks where mistakes are cheap to fix, like internal drafts, before automating anything customer-facing.
- Decide who reviews what. Salesforce itself recommends that a person check each model response before sharing it, for most use cases 6.
- Measure the effect on reps' work, beyond adoption. Gartner predicts that by 2028, fewer than 40% of sellers will report that AI agents improved their productivity 7.
The risks worth watching
The most common one is inaccuracy. In McKinsey's 2025 survey, nearly a third of respondents reported consequences from inaccurate AI output 2. NIST describes the problem as confabulation, the confident production of erroneous or false content 3.
In sales, a mistake costs trust. A wrong figure in a proposal or a misspelled name in an email can wear down months of relationship. And buyers do check, since 69% of B2B buyers turn to sales reps to validate the insights they get from AI, according to Gartner 8.
Where relationship intelligence changes how sales teams use AI
Almost every application above depends on context that rarely sits in the CRM, like who actually decides, what was promised and what changed in the account. Without it, AI writes fluently about things it doesn't know. Relationship intelligence starts from the memory of the relationship, with a source and a date on every fact, and only then uses AI to summarize, prioritize and draft.
In ARC, the team logs conversations as voice notes, screenshots, messages or recorded meetings, and three engines work on top of that memory. The Opportunity Radar links outside signals to accounts, the Infinite Funnel finds relationships with no next step, and the Content Hub prepares proposals and briefings with the source of every passage. Every suggestion goes through a person before it goes out.
FAQ
What are the main uses of AI in sales?
Summarizing conversations and records, preparing meetings, drafting emails, prioritizing opportunities, suggesting next steps, researching accounts, responding to RFPs and tenders, tracking outside signals, supporting coaching and keeping relationships alive outside the pipeline.
Will AI replace salespeople?
The most mature uses take repetitive work off reps and leave the decisions with them. B2B buyers still turn to reps to validate what AI tells them 8.
Do I need a new CRM to use AI in sales?
Not necessarily. Many CRMs already include AI features, and specialized tools can work alongside your existing CRM. What decides the outcome is the quality and completeness of the data the AI will use.
Sources
- State of Sales, 6th edition (2024): sales and AI statisticsSalesforce
- The state of AI in 2025: Agents, innovation, and transformationMcKinsey & Company
- NIST AI 600-1, Artificial Intelligence Risk Management Framework: Generative AI Profile (2024)National Institute of Standards and Technology
- Unlocking profitable B2B growth through gen AI (2025)McKinsey & Company
- Gartner Survey Finds Sales Organizations That Provide AI-Enabled Next Best Actions Are 2.6 Times More Likely to Achieve Commercial Growth (2026)Gartner
- Trust Layer, Agentforce Developer GuideSalesforce
- Gartner Predicts by 2028 AI Agents Will Outnumber Sellers by 10X (2025)Gartner
- Gartner Survey Finds 69% of B2B Buyers Turn to Sales Reps to Validate AI-Generated Insights (2026)Gartner
- Sales AI: Agentforce and Einstein for Sales CloudSalesforce
See ARC with your own team.
The guided demo shows, in about a minute, how one conversation becomes memory and a next step.