The paid version provides access to more up-to-date information and allows you to create company and activity-specific AI tools to streamline and refine your process. However, these preparation materials should themselves be further refined and supplemented with information where appropriate, to ensure that they can be used to their full advantage in the customer encounter itself.

Artificial intelligence to support sales in creating conversation summaries

I tend to take notes on each client encounter, so that the next time I meet the person I remember our previous conversations as well as possible. And of course, many organisations also require these notes to be recorded in the CRM anyway, so at least some sort of note has to be made anyway.

Often, however, these notes are a bit confusing, as the topic may change during the discussion and the theme of the discussion may change. And heaven help us if this information has to be interpreted by someone else who was not at the meeting - it won't make any difference.

In this case, the notes, at least in that form, will no longer really refresh the memory when the next encounter takes place, but may be confusing, or at least take an insane amount of time to decipher. However, AI has become quite good at supporting in such situations.

In practice, the discussions are summarised as follows:

Step 1: Transfer your notes to electronic format

The notes need to be in a format that the AI can understand, so that they can be summarised. So if your notes are already in Word, for example, there is no need for this step.

Step 2: Anonymise the text

By default, many AI solutions also share the data you enter with other users. Because of this remove any reference to a company, person or other identifiable element from the raw text of the memo., so that discussions held in confidence do not fall into the wrong hands.

This is unlikely to be necessary when using business-specific AI tools. However, please check with your organisation's AI manager before you act!

Step 3: Make your text available to AI

Enter the anonymised text as a file or text to the AI and ask for a summary of the text. A good prompt or request to the AI could be, for example, ”Condense the text below into a note in CRM” or ”Condense the attached text into a note in CRM”.

Step 4: Check the AI's summary

In general, AI can make good summaries, but sometimes it tends to draw conclusions that are too straightforward. That's why the summary still needs to be checked to make sure that no false assumptions or other ”hallucinations” have been made.”.

At this stage, also include any identifying information for internal use if you removed it earlier when anonymising the text.

Step 5: Export the data

With the memo finally in order after review and completion, it's time to transfer the data to the recovery in the right place. The most typical storage locations are the CRM system or other place where you have agreed to share this type of information.

The list is long, but in reality it takes a few practice sessions, a few minutes at most. No special paid tool is needed, e.g. a free of charge ChatGPT is able to summarise practically as well as the debit side.

However, if your company had small investment funds available for technology development, for example. Copilot with a paid company-specific version it is possible to skip most of these steps during remote meetings, as it creates notes directly from Team discussions.

Artificial intelligence to support sales analytics

After using AI for a while, you as a salesperson will probably notice that data on these encounters and opportunities has started to accumulate quite a bit. There are memos, chalk lists, preparation materials, you name it. Now is the perfect time to put this data to use.

A word of warning: this is the only one of the habits that may require a little more work in the midst of the everyday hustle and bustle. So it's worth setting aside a little time for your first experiment, so that you don't get caught up in the rush and lose enthusiasm.

Step 1: Think about the question you want answered

It all starts with a question. For example, have you considered what kind of customers are really buying from you? And why some customers have chosen another partner? Or do you dread the thought that you've inadvertently passed up a potential prospect?

So choose the question for which getting the most objective answer is likely to improve the effectiveness of your work. For example, let's use the question of buying customers as an example.

Step 2: Collect the data you need

It's time to open a CRM or some other tool where you keep track of all those customers you think are buying. So take all the information you need, say in Excel or Word, about the companies that have bought from you or, more broadly, from you in, say, the last two years.

This information can usually be extracted in bulk from modern CRM systems. If necessary, you can also enter the data manually or ask the system administrator to create a list of these companies - especially if there is a lot of data and you cannot get it out of the system yourself.

Step 3: Ask AI for help

Now is the time to open the AI tool and ask for help. The one used in the example In the paid version of ChatGPT you can attach the list you've collected as a file directly to the AI for interpretation, whereas with the free version you'll probably have to phrase the question slightly differently and list the businesses underneath separately manually.

Please note that if you do not have a customised environment created for your company, the data should be anonymised, if necessary, before being passed on to AI. However, if it's just a list of companies, there shouldn't be any problems.

Below is an example, using a fictitious list of six companies as an illustration. Each of the companies employs at least 100 people and its main activity according to the TOL2008 classification is software design and manufacturing.