Ask the Expert: When Should You Use AI in the Sales Process?

Using AI in the Sales Process: Jeff’s Buckets

The first part of the discussion with Jeff focused on the buckets he breaks AI into when talking about it as a sales tool:

• Prospecting
• CNA (customer needs assessment) due diligence
• Creative and ad setup

AI and Prospecting

Decades ago, local media sellers might have driven around their communities to gather intelligence about potential prospects. Doing that today would be difficult, so AI can really be a driver for this.

The prompts matter, of course, in terms of what you get. Jeff advised that you have to keep refining these for the desired result.

Some prompts for prospecting include:

• Create a list of businesses in (ZIP code) that are at least three years old, have between $2,000,000 and $10,000,000 in gross annual sales and primarily serve the B2C sector.
• Provide me with two advertising and marketing decision-makers from each company, including their name, address, phone number and email wherever possible.
• Create an outreach email template that I can use for these prospects. Be aware of header length, as smartphone screens may truncate long header copy. Include a customizable Valid Business Reason within the template. The email template must be able to be read by the recipient in under 15 seconds.

AI and CNAs

Once you’ve defined a good prospect through your research and booked a meeting, AI can assist with your prep. It’s good at summarizing information that’s specific to an industry, so you can pick out some high notes that show you understand their business.

Jeff suggested the example of asking AI to “teach him about the category of orthopedic practices.” It’s a very specialized area of health care, and gathering information about the industry offers some brushstrokes.

Then, you can get more specific, Jeff explained, by feeding the AI engine the business’s URL and location. The output could help you identify things for this practice in this area.

Traditional research still works, but it’s takes more time than AI, which can bring you up to speed quickly.

Another way to use AI pre-CNA is to define a budget range. Most companies won’t disclose what they have to spend. You can guess and lay out lots of options, but AI may be able to help.

You need to feed it information such as annual revenue, number of employees, locations, and industry averages. It will be a range and may vary due to seasonality, but it can serve as a starting point. Jeff noted you could also compare this to other resources like those from the RAB.

AI and Creative and Ad Setup

The third bucket Jeff spoke about is how you can use AI to help with creative scripting. He recently worked with a salesperson who needed a script for a CTV ad as part of a branding campaign.

They had only a few hours to turn something around for the advertiser, a nonprofit, but Jeff was able to produce something in under one hour using AI. There were unique criteria, as they wanted to target a ZIP code where donations had been very small compared to others.

He accomplished this through a series of prompts:

1. Break down the dynamics of this ZIP code.
2. Tell me about the nuances of this community and what makes it different compared to those around it.

From the first two questions, he was able to establish knowledge of the area. Then, he took what he learned and the information about the advertiser and asked AI to draft a script. He continued refining it through additional prompts until it was the correct length.

Jeff championed this approach to using AI with creative. It needs to be a multi-step approach. The final script from AI also required a human review. Tweaks were made to personalize it and iron out any parts that felt too robotic. The essence was there, as the copy detailed community events and landmarks, and the client was very pleased.

Letting AI be your sole script or ad copywriter is not advised. Even with perfect prompts, everything still needs a human touch! However, this exercise saves a lot of time and allows you as a seller to “own the creative” and be a source of ideas.

Jeff cited SEM (search engine marketing) keyword research as another way to use AI. It can develop a list and categorize them into ad groups. It still needs a final review, but it’s much easier than doing it from scratch.

AI and the Sales Process: When to Be Cautious

With all the opportunities that AI brings to the sales process, Jeff also offered some wisdom on when you can’t completely rely on the technology.

Don’t take any output as gospel.

You still need to review and validate it. Human review is always necessary.

Avoid relying on it solely to recommend tactics.

There are some use cases for this now, but it’s more about an algorithm analyzing past performance than a full AI model. It may offer some good suggestions, but you don’t start CNAs talking about tactics. You begin with uncovering challenges and finding solutions.

Any tools that claim to do this based on industry will also have gaps. There are varying ad goals for any industry. For example, roofing companies run ads to acquire customers who have roof damage. They also have campaigns in search of new employees. These are entirely different things.

Better Prompts, Better Results

Jeff relayed that the more you improve your prompting, the better the output will be. The more specific, the better the results. You may have to go through a series of prompts, but this allows you to learn, too, about what to expect from AI and its weaknesses.

When the prompt is aware of the campaign’s intent, such as for SEM, you’ll get the most relevant keywords.