AI can help with MSP sales prospecting, but I wouldn’t spend ten minutes asking an LLM to research every company before making a call.
Instead, use AI to remove the administrative work that slows prospecting: cleaning lists, extracting data from files, standardizing fields, and organizing information into a usable calling list.
The hidden cost of over-researching prospects
Where AI fits into the sales process matters.
According to our benchmark, it takes about 300 dials to generate one qualified first-time appointment (FTA). A caller can complete those 300 dials in about five working days when the time between calls stays under 90 seconds.
Now add ten minutes of research before every call.
Those 300 dials require another 2,550 minutes, or 42.5 hours, on top of the existing calling process. Your five-day path to one qualified FTA becomes roughly 82.5 hours of work, or more than ten working days.
You haven’t generated another appointment. You’ve doubled the labor required to get one.
If your fully loaded BDR salary is $1,500 per week, each FTA now costs more than $3,000. If your team closes one out of every three FTAs, the acquisition cost exceeds $9,000. You will also generate roughly half as many appointments each year.
Research has a place in the sales process, but not at the top of the funnel.
Don’t research prospects who haven’t engaged
Most companies on a prospecting list will never answer the phone. Many won’t respond to email. In addition, 30 percent or more of the data from a provider may be outdated or inaccurate.
An LLM cannot automatically access everything you can. Some tools have no web access, while paywalled or login-protected sources remain unavailable unless connected through approved integrations.
Do enough targeting to confirm a company belongs on your list, then start prospecting.
Once a prospect responds and meets your MSP’s qualification criteria, deeper research becomes worthwhile.
Give AI the administrative work
West McDonald of goWest.ai offers a useful rule for deciding where AI belongs:
“If you wanna know the best way to use AI in your business, start with the work.”
His advice is simple. Identify the tasks that take the most time and create the most frustration. Understand the process first, then decide where AI can help.
That’s exactly how I think about prospecting data.
Suppose you want to call law firms in your market. You search Google Maps, save the results as a PDF, and collect business names, addresses, phone numbers, websites, and other details.
If your approved AI tool can read PDFs, upload the file and ask it to extract the information into defined columns. Instruct it to leave unknown fields blank. Export the results to CSV, review the data, and you have the foundation of a calling list.
The same process works with association directories, event lists, legacy spreadsheets, and other approved prospect data sources. AI can standardize phone numbers, separate address fields, identify missing information, and flag duplicate records.
Use AI to organize data, not invent it
Brandon Borden ran into this challenge while trying to identify MSPs at scale for Fixify.
He explained it clearly:
“For us to burn through a bajillion tokens and still potentially get it wrong was not worth my time or the org’s time.”
His team wanted to identify prospects, determine which companies matched their ideal customer profile (ICP), and add technographic data. That work requires reliable source data and industry-specific knowledge.
Asking an LLM to fill in missing information did not solve the problem.
That’s where I draw the line with prospecting.
AI excels at restructuring and organizing information you already have. It becomes less reliable when it must provide data or market knowledge it was never given.
Missing data should remain missing until you obtain it from a trusted source.
Create a free prospecting list with Google Maps and AI
- Pick one market and one geography. Search for a business type that fits your MSP’s target market, such as “law firms near me” or “accounting firms in Detroit.”
- Collect the Google Maps results. Save or print the results to a PDF. Focus on gathering records, not conducting extensive research.
- Upload the file to an AI tool that can read documents. Use a tool approved for the data you are uploading.
- Tell the AI exactly what to extract. Request fields such as company name, address, city, state, ZIP code, phone number, website, and business category.
- Tell it not to guess. Add a simple instruction: “If the source does not contain the information, leave the field blank.”
- Request CSV output. Format the extracted records so they can be imported into a spreadsheet or CRM.
- Verify the results. Review a sample for duplicates, formatting issues, shifted columns, and data that did not exist in the source file.
- Clean the list before importing. Remove duplicates and records outside your target market or geography.
- Import the list and start calling. The list does not need to be perfect. Early conversations often provide more useful insights than ten minutes of speculative research. If prospects never answer the phone, additional research will not change that.
Focus AI on what it does best
AI doesn’t need to know everything about your prospects to add value. Its biggest strength is handling repetitive work that keeps salespeople from making the next call.
If you’re evaluating where AI fits into your MSP sales prospecting process, consider scheduling time with Carrie Richardson at Fox & Crow Group.
Photo: nampix / Shutterstock
This post originally appeared on Smarter MSP.