Why use AI instead of paying a data scraper?
Hiring someone to scrape business contact data used to be a real option, and it cost money. AI tools like Gemini and ChatGPT do the same job for free and return results in seconds. You get business names, websites, phone numbers, emails, physical addresses, and key decision makers without spending hours searching the internet or paying a third party.
Which AI tool should you use?
Either Google Gemini or ChatGPT will do the job. Greg uses Gemini because it connects directly to Google Sheets, which means you can export your results to a spreadsheet in one click. ChatGPT produces the same type of output. There is no meaningful loyalty required here: use whichever one you are comfortable with.
How do you write a prompt that actually works?
The more specific your prompt, the better the results. A useful prompt should include:
- The types of businesses you want to target (for example: builders, facilities management companies, body corporates, strata managers, landscapers)
- A geographic area and radius (for example: within a 20-kilometre radius of Newcastle, New South Wales)
- The exact columns you want returned (business name, website, phone number, email, industry, physical address, key decision maker)
- Any social media or professional profiles you want included (Instagram, Facebook page, LinkedIn profile link)
- The output format you prefer (a structured table, CSV format, or both)
If the first result comes back with fewer businesses than you expected, simply follow up by telling the AI the list was short and asking it to return as many businesses as possible. It will keep going.
How do you get the results into a spreadsheet?
Ask for the output in CSV format. CSV is a file format that loads directly into Excel or Google Sheets. In Gemini, there is an option to export the table straight to Google Sheets without any copying. If you use ChatGPT, you can copy the CSV code and paste it into your spreadsheet manually. If you do not need a spreadsheet at all, just leave the CSV instruction out of your prompt and ask for the information in plain list format instead.
How accurate is the data?
The data is not always current or complete. Contact details go out of date, decision makers change roles, and some fields will be missing. Greg points out this is the same problem you face when you look up a business on their own website: the manager listed might have moved on. You should treat the output as a strong starting point that still needs a quick check before you pick up the phone or send an email. The time saving compared to manual research is still significant, even if you need to verify a few entries.
Can you tailor this to a specific trade or industry?
Yes. If you are a heavy diesel mechanic, for example, you would adjust the business types in your prompt to construction companies, earth-moving contractors, and mining support businesses. Any industry vertical works. The principle is the same: describe exactly who you are looking for, where they are located, and what information you need, and the AI will build the list for you.
