Should you just give the whole team ChatGPT?
Rolling a general AI out across the firm scales access to a chatbot, not your way of working. What ChatGPT, Gemini and Copilot do for a land team, and what they don't.
4 min

Should you just give the whole team ChatGPT?
It is a fair question, and the honest answer starts here: the general AI tools are good. ChatGPT, Gemini, Microsoft Copilot, Claude. If your team is not using one yet, that is the first thing to fix, and it costs very little to start. We run a free Claude-for-Property course precisely because most land firms are under-using them.
But "should we use a general AI" and "does giving everyone a general AI make the firm faster at winning sites" are different questions. The first is easy. The second is the one that matters, and the answer is more interesting.
Rolling a general AI out across the firm scales access to a chatbot. It does not scale your firm's way of working. Three things stay exactly where they were.
Everyone still starts from zero
A general AI knows the internet. It does not know how your firm prices a promotion agreement, the comps you trust, or the constraint you always check first. So every person on the team re-explains the firm to it, every time they open a window. You have not given the team your methodology. You have given ten people the same clever intern who has never seen one of your deals.
Harold is trained on your firm's deals. It prices the way you price and screens against the constraints you care about, the same way for everyone, because it learned them once.
Nothing holds the deal
A land deal runs for months, sometimes years. A chat thread runs until someone closes the tab. No one's chat history knows the landowner went quiet in May, that the option expires in March, or that highways is still outstanding. The deal still lives in people's heads and a spreadsheet, which is exactly where deals go quiet and slip.
Harold holds the state of every live site across the whole pipeline, and flags what is about to slip before it does.
Nothing does the work where the work happens
This is the one people underestimate. A general AI drafts. Someone still copies it out, formats it, pulls the addresses, runs the merge, attaches the file and sends it. You have made the thinking faster and left the doing exactly where it was, for every person, on every task.
Harold runs the work inside the tools you already use. Screening against your brief, the appraisal, the letter pipeline end to end. Not a draft to action. The work, done.
"But Copilot is already in our Microsoft"
Fair, and Copilot is the closest of the lot on one point: it lives inside Outlook, Word and Excel, so it is not stuck in a separate window. That is real. But it is still a general assistant sitting in your apps. It does not know your deals, it does not hold the state of a live land transaction, and it does not run the land workflow. Closer on the plumbing, same gap on the job.
So which do you buy?
Both, usually. They do different jobs.
A general AI makes each person a little faster at their own tasks. It is worth doing, cheap to start, and we will happily show your team how. Harold makes the firm faster at winning sites, because it knows your deals, holds every live site, and does the analyst-and-assistant work from end to end. One is a tool you hand out. The other is someone the firm gains.

Article written by
Sam Sykes


