AI in sales:better prepared for every call,faster to the proposal

AI in sales takes the preparation and follow-up off your team’s plate: it gets you ready for customer calls, summarises email threads, drafts proposals and follow-ups and turns rough notes into a clean record. The conversation itself, the price and every commitment stay with a person. If you work with customer data, you need a business account with a data processing agreement, and marketing emails remain subject to competition law whether a person or an AI wrote them.

Nikolai Schöbel and Jeremias Burger

Co-founders of Scalableloops GmbH. Nikolai Schöbel leads online marketing and AI strategy, Jeremias Burger the AI architecture. Both build AI systems and train teams on them in their own agency work.

Which AI tools are suitable for sales?

You will come across three kinds of tools. Which one fits depends on where your team works today and where your customer data lives.

General AI chats such as ChatGPT, Claude or Gemini are ready to use and good for drafts, role plays and tidying up notes. They only know your customers as far as you tell them, though. For customer data they are only an option with a business account, which the data protection section covers.

Assistants built into office software such as Microsoft Copilot work where your emails and meetings already are. Microsoft lists summarising email threads in Outlook and summarising meetings in Teams. Our article on the AI assistant for business explains the different types.

AI built into CRM systems works with your customer records. For Copilot in Dynamics 365 Sales, Microsoft describes summaries of leads and opportunities, preparation for upcoming meetings and help composing emails. HubSpot offers a Prospecting Agent that monitors buying signals and prepares personalised outreach, and Salesforce offers an Agentforce agent for first contact with prospects. Features like these go beyond an assistant because they complete several steps on their own. Our article What is an AI agent? explains what that means.

How do you write proposals and follow-up emails with AI?

The quality of an AI draft depends almost entirely on the prompt. A vague request such as “Write a follow-up email” produces a generic text that could go to anyone and therefore speaks to no one. Prompts work better with four building blocks:

1. 01

   01

   Role

   Who is writing? For example: “You are writing as the sales manager of a small engineering firm, factual and friendly.”

2. 02

   02

   Context

   What has happened? Paste in call notes, the customer’s enquiry and your proposal template.

3. 03

   03

   Goal

   What should the customer do after reading? For example, confirm a meeting or answer an open question.

4. 04

   04

   Format and limits

   Length, structure and what must not appear: “No more than 120 words, no prices, no discounts, no sales clichés.”

Can you send AI-personalised marketing emails to new contacts?

AI makes it easy to write hundreds of personal-sounding emails to companies you have never dealt with. The legal position stays the same. In Germany, Section 7(2) no. 2 of the Act against Unfair Competition (UWG) treats advertising by electronic mail without the recipient’s prior express consent as an unreasonable nuisance. The law does not distinguish between private individuals and businesses.

There is an exception for existing customers under Section 7(3) UWG: you may email customers whose address you obtained in connection with a sale about your own similar goods or services, as long as they have not objected and are clearly told, when the address is collected and in every email, that they can object at any time.

Cold calls follow a different standard: towards businesses, Section 7(2) no. 1 UWG requires at least presumed consent, towards consumers prior express consent. AI helps you prepare calls and emails better. Who you are allowed to contact is still decided by the law.

If you use an AI chat on your website to advise prospects, the disclosure obligation in Article 50 of the EU AI Act has applied since 2 August 2026. What belongs in the first message is covered in our article on AI disclosure for chatbots. How AI handles enquiries after the sale is the subject of Use AI for customer service.

Where are the limits of AI in sales?

AI in sales rarely fails because of the technology. It fails because of the wrong expectations. Four mistakes are typical.

Invented facts. Language models fill gaps with plausible-sounding details, such as a wrong headcount or a project that never existed. In front of the customer, that shows immediately.

Interchangeable texts. Without your notes and your tone, the AI writes like everyone else. Customers recognise mass emails even when their own name is in them.

Commitments without cover. A draft can quickly promise a delivery time or a service your business does not offer in that form. Anything that commits your business is checked by a person.

No training. Article 4 of the EU AI Act has applied since 2 February 2025 to organisations that use AI systems. Since 27 July 2026 it applies in amended form: you have to take measures that support the development of AI literacy, but you do not have to guarantee a specific level for each person. No certificate is required. For sales, that means anyone using AI should know which data may go in, where the tool tends to get things wrong and who signs off. Details are in our article on the AI literacy obligation under Article 4.