AI for sales teams: prep faster, follow up consistently

Selling depends on preparation, follow-through and trust. AI can help with the first two. It is not a substitute for the third, and buyers are getting better at recognizing message templates written by a model.

Where AI helps sales teams

Preparation. Before a call, gather what you can about the company: what they sell, recent news, leadership changes, likely priorities. Tools that search the web and show sources, such as Perplexity, are good at this. Open the sources for anything you plan to mention. See Perplexity vs Google for how these differ from classic search.

Notes and CRM hygiene. Most CRMs are only as good as the notes reps have time to enter. Record or take rough notes, then ask for a structured entry: pain points, decision makers, timeline, budget signals, next step with date. It gets the record done while the call is fresh.

Follow-up. The recap email after a call is the most valuable and the most often skipped. Paste in your notes and get a draft that restates what you heard and proposes the next step. The Gmail guide shows how to build inbox workflows that draft replies and flag items needing action.

Proposals. Give it your template, the call notes and the price list, and it assembles a first draft. You verify pricing, terms and scope.

Pipeline reviews. Export the CRM and ask for deals that moved, stalled or have no next step. It lets a manager prepare for a weekly meeting in minutes.

Where AI does not help

• Facts about the prospect. It may attribute a funding round, a product or a quote to the wrong company. This is a hallucination. Verify anything you would say out loud.
• Relationships. Bulk personalization at scale reads as bulk.
• Forecasting. It can summarize a pipeline but does not know which deals are real.
• Pricing and discounts. Keep these under your approval process.

Risks specific to sales

Customer and prospect data. CRMs hold personal contact data and confidential deal terms. Uploading exports into a general AI tool may conflict with your company policy, customer agreements or privacy law. Read is it safe to put company data in ChatGPT.

Recorded calls. If you record or transcribe calls, recording consent laws vary by state and country. Confirm your practice with your legal team.

Commitments in writing. An AI-drafted email or proposal can promise a feature, date or price you cannot deliver. Read it as if a customer will hold you to it.

Outreach rules. Email and messaging regulations apply regardless of who wrote the message. Follow your company’s compliance guidance.

Starter prompts

Account brief:

Research [company] and write a one-page pre-call brief: what
they do, who their customers are, three recent developments
(with source links), and two likely priorities related to
[our product category]. Mark anything you could not confirm
from a source.

Call recap:

From my call notes below, write a follow-up email under 150 words:
thank them, restate the two main problems they described in their
words, propose one next step with a date, and attach nothing that
was not discussed. Notes: [paste]

Your first week

• Day 1: Pick five upcoming calls and produce briefs. Verify one fact per brief.
• Day 2: Convert notes from your last five calls into CRM entries and compare to what is there.
• Day 3: Send recap emails after every call, edited from drafts.
• Day 4: Build a one-page proposal from your template and check every figure.
• Day 5: Have your manager run the pipeline summary prompt and compare with what the team says.

A worked example: the 15-minute pre-call

You have a call in an hour with the operations director at a mid-sized manufacturer. Ask for a brief with source links: what they make, recent news, leadership changes, and two likely operational pain points relevant to what you sell. Open two of the links to confirm. Then ask for five discovery questions tailored to those pain points, phrased the way you would say them, not as a script.

On the call you sound prepared without sounding rehearsed, because you verified the facts you use. Afterward, paste your notes and get a recap email and CRM entry. The whole loop replaces an hour of scattered tabs.

Common mistakes to avoid

• Mass-personalized outreach. If ten thousand messages say “I loved your recent post”, buyers notice.
• Quoting facts you did not verify. Getting a prospect’s own business wrong ends the call.
• Letting a draft promise a feature. Read it as a contract.
• Uploading the full CRM export into an unapproved tool.

What the sources say

BLS data on sales managers and sales representatives, and the FCC ruling that affects AI-assisted outreach by phone.

• The Bureau of Labor Statistics projects employment of sales managers to grow 4 percent from 2025 to 2035, about as fast as the average for all occupations.
• The Bureau of Labor Statistics projects little or no change in overall employment of wholesale and manufacturing sales representatives from 2025 to 2035.
• The Bureau of Labor Statistics says online sales of wholesale and manufacturing products are mostly expected to complement, rather than replace, face-to-face selling.
• The Bureau of Labor Statistics says advances in AI and chat bots may limit employment growth for wholesale and manufacturing sales representatives.
• The Federal Communications Commission ruled that calls made with AI generated voices are “artificial” under the Telephone Consumer Protection Act.

To go further with your team

Learn AI Fast is a two-day in-person class in Boston for non-developers. Everyone works through the same exercises, which suits sales teams that want a shared starting point.

Sources

• U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Sales Managers: “projected to grow 4 percent from 2025 to 2035, about as fast as the average for all occupations”
• U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Wholesale and Manufacturing Sales Representatives: “projected to show little or no change from 2025 to 2035”
• U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Wholesale and Manufacturing Sales Representatives: “These online sales are mostly expected to complement, rather than replace, face-to-face selling”
• U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Wholesale and Manufacturing Sales Representatives: “integration with automation technologies, such as chat bots, may limit employment growth for sales representatives”
• Federal Communications Commission, FCC Makes AI-Generated Voices in Robocalls Illegal (February 8, 2024): “calls made with AI-generated voices are “artificial” under the Telephone Consumer Protection Act”