What was the result?
self-reportedEric Bramlett, broker-owner of Bramlett Partners in Austin, Texas, writes that over the last year his brokerage replaced roughly $2,500 a month in off-the-shelf software subscriptions with internal tools built using Claude Code.
When the Austin real estate market turned, Eric Bramlett audited what the brokerage was paying for. In a May 13, 2026 article he wrote for REALTOR News, he says overlapping software was a quiet drain on budget and agent attention. Over the last year, the team replaced roughly $2,500 a month in off-the-shelf subscriptions with internal tools built in Claude Code. He writes that they are not a technology company, and that the question has shifted from what to buy to whether they should build it first. He does not name which subscriptions were replaced. Other cuts in the same article, including Google Ads and farm mail, were not done with AI and are not part of this result.
Who did this?
How do I try it?
- Editorial adaptation: list every software subscription with its monthly cost, who uses it, and the one or two features people actually use.
- For each one, ask whether it is producing closed deals, whether someone is actively managing it, and whether cutting it would leave a real gap.
- Pick one paid tool where you only use a simple feature, such as a deal tracker or a checklist, and describe that feature list to Claude Code.
- Run the new tool alongside the paid one for a few weeks on non-confidential data, and keep client data out until you have approved where it is stored.
- Cancel the old subscription only after the team has stopped opening it, and review the build yourself before anyone relies on it.
Editorial adaptation, not Eric Bramlett's exact prompt. I run a small real estate brokerage. We pay for a tool whose only features we actually use are: [feature list]. Our team of [N] people uses it to [job]. Build the smallest version that does exactly these features, as [a single web page that saves to a CSV file, or a Google Sheets script], with no logins or payments. Before writing code, list what you will build, what you will not, and any client-data or privacy risks I should check. Then build it, explain how to run it in plain English, and give me 5 tests to try. I will review it before anyone on the team uses it or before I cancel the old tool.
Make it run every week
A play works once. Turn it into a routine in three small steps:
- Run it once and check the result yourself.
- Give your AI a schedule. Tell it when to run this play (every week, or every month) and to message you with what it found before you act.
- Keep a log. One line per run: what was checked, anything flagged, what you did.
What should I watch out for?
Self-reported in an article Bramlett wrote for REALTOR News on May 13, 2026. The AI-specific figure is roughly $2,500 a month, over the last year as of that date; do not treat the article headline of more than $40,000 in monthly expenses as an AI result, because most of that is Google Ads and farm mail cuts. He does not name which subscriptions were replaced, so this is not a recipe for a specific tool. Keep client and deal data out of a new build until you have approved where it lives. Build it with Page: a private subscriptions audit page that lists each tool, its cost, and whether you still need it.
Did it work for you?
PageMotor keeps this site free.