Skip to main content

,

ChatGPT prompts for credit managers

Staying ahead often means embracing innovative tools and technologies. One such tool gaining traction is ChatGPT, an advanced language model that can assist credit managers in various aspects of their work.

Staying ahead often means embracing innovative tools and technologies. One such tool gaining traction is ChatGPT, an advanced language model that can assist credit managers in various aspects of their work.

Why it matters: To get the most out of ChatGPT, credit managers need to know exactly what to ask.

Carefully navigating your conversation with ChatGPT is vital for efficient workflow and accuracy. NACM has gathered insights from credit professionals on what they ask ChatGPT to get the information they need.

Customer research prompts

A recent eNews poll revealed that 23% of credit professionals use ChatGPT to research customer information. A process that once took weeks or even months can be shortened to less than a week.

Chantal Rousseau, CCP, corporate credit director at MPG Canada (Brossard, QC), uses ChatGPT to research customer information such as finding the relation between two entities or the owner of a corporation. 

“Depending on the case, I will ask the following variations, such as, ‘Can you tell me what the relationship is between X and Y?’ or ‘How are X and Y related?’ or ‘Can you find any proof of a relationship between X and Y,’” Rousseau said.

Below, ChatGPT has outlined a handful of specific prompts credit professionals can use for various aspects of the credit investigation process:

  1. Financial analysis:

“Can you analyze the financial stability of Company XYZ based on its recent annual reports?” or “What are the key financial metrics I should consider when assessing the creditworthiness of Company ABC?” or “Compare the financial performance of Company DEF to its industry peers over the past three years.”

  1. Industry insights:

“What are the current trends and challenges in the [industry/niche] that Company LMN operates in?” or “Provide an overview of the competitive landscape for Company PQR’s industry,” or “What regulatory changes or market shifts could impact the credit risk of businesses in the [specific sector]?”

  1. Company background and reputation:

“Can you provide a summary of any recent news or controversies surrounding Company UVW?” or “What is the reputation of Company EFG within its industry and among its customers?” or “Retrieve any information on significant management changes or legal issues for Company HIJ.”

  1. Risk assessment:

“Identify any red flags or warning signs indicating potential credit risk for Company ZAB,” or “Provide insights into the credit risk associated with extending a line of credit to Company CDE,” or “Assess the likelihood of default for Company FGH based on its historical performance and market conditions.”

  1. Customer behavior and payment history:

“Retrieve any information on late payments or defaults by Company IJK in its previous transactions,” or “Analyze the payment behavior and credit history of Company OPQ with its suppliers,” or “Compare the payment patterns of Company XYZ with industry benchmarks to identify any anomalies.”

Writing prompts

ChatGPT generates high-quality content by providing ideas, drafts and improvement suggestions for writing.

To write emails, contracts or dunning letters, Elliott Jenneman, CCE, regional credit manager at McNeilus Steel Inc. (Dodge Center, MN), uses the following language: ‘Please assist in helping me sound more (insert adjective like professional, clear or concise) in this (type of document; email/letter/communication).’

Although not officially implemented in the workplace, Martin Smith, CCE, CICP, credit manager at Ash Grove Cement Company (Sumterville, FL), uses ChatGPT to generate collection letters. He’d type, ‘Write me a collection letter with a slightly aggressive or very low-key tone.’

“I also use a couple of keywords and then it’ll generate the letter, leaving the customer’s name out of it because I don’t want to leave a trail of their information,” Smith said.

Credit managers should use plain, easy-to-understand language and limit corporate jargon for cleaner responses from ChatGPT. Jenneman asks for multiple drafts by saying, ‘Provide four drafts for me to choose from.’ After picking the one he likes best, he combines sentences and word choices to write his own letter.

Reviewing prompts

One of ChatGPT’s most desirable qualities is that it can offer suggestions, identify errors, provide alternative wording or give general critiques based on the content it analyzes.

You can ask it, ‘Rewrite this collection email to be more professional,’ or ‘Review this contract language for pay-when-paid clauses.’

Sometimes the return is not as clear as you would wish and you have to communicate changes, said Michelle Kelly, CCE, CCRA, CICP, senior credit manager at Mansfield Oil Company of Gainesville, Inc. (Gainesville, GA). She has prepared a list of prompts to review her work:

  • Make this more concise
  • Condense into two-three paragraphs
  • Make into bullet points
  • Use this to create a business plan
  • Remove contractions
  • Use a more professional tone
  • Use a kinder tone
  • Revise to be less formal
  • Remove “esteemed” and replace with “respected”
  • Add this specific data
  • Remove this specific data

“The software occasionally struggles with context and may misinterpret or have difficulty grasping the subject matter,” Kelly said. “Therefore, it’s crucial to review the responses for accuracy.”

Using keywords

Although keywords do not generate detailed ChatGPT responses, they can lead you in the right direction and save you time from crafting a specific prompt. Kelly has a set of standard keywords she uses in ChatGPT:

  • Rewrite
  • Create
  • Review
  • Develop
  • Create an outline, step-by-step process
  • Task with an example (mathematical, strategy, business plan etc.)

To research customers on ChatGPT, Smith uses keywords along with the company’s name to look for public filings such as news reports.

Although there are countless keywords used in the B2B credit profession, here are 10 keywords credit managers can use when interacting with ChatGPT:

  1. Trade credit
  2. Credit terms
  3. Credit application
  4. Credit risk assessment
  5. Credit analysis
  6. Payment terms
  7. Credit policy
  8. Credit limit determination
  9. Credit evaluation
  10. Credit scoring models

Asking questions 

Rather than using keywords or short phrases to generate responses, try asking ChatGPT questions to get the answers you need. For example, Kelly frames her questions as such, ‘What are some best practices for managing…’ or ‘How does …’

Most of the time, ChatGPT prompts vary on a case-by-case basis. Credit professionals must ask more specific questions to generate relevant responses. Rousseau asks more complex, detailed questions like, ‘If you are being told XYZ, and you are a credit manager, how would you respond?’

Depending on the answer you get from ChatGPT, you can ask supplemental questions to get the right answer for you. “If something is not clear, I would ask to clarify what XYZ means,” Rousseau said. “Or if the answer did not address a portion of the complaint, I would want to revisit that part. It would really depend on how much I like the answer or not as well.”

The bottom line: To maximize the benefits of ChatGPT, credit professionals should continuously refine their questioning techniques and adapt to the responses received.

Jamilex Gotay, editorial associate

Jamilex Gotay, a Towson University alum, holds a B.S. in English. Her creative writing background fuels her success as a writer, journalist and award-winning poet. Fluent in English and Spanish, with intermediate French skills, she’s passionate about travel and forging connections. When not crafting her latest B2B credit story, she enjoys quality time with loved ones, outdoor pursuits and creative activities.