Why brokerages waste money on AI tools agents don't use
Brokerages are buying AI platforms agents barely touch. The ones seeing real ROI aren't buying more tools. They're embedding AI into 2-3 workflows agents actually care about.
Quick Answer: Most brokerages have a tech stack full of AI-branded tools agents never open. The ones getting measurable ROI aren't adding more features. They're wired AI into 2-3 money-critical workflows (lead follow-up, listing prep, negotiation) where agents already spend their time.
Darryl Davis published an analysis in Inman in June 2026 that quantified what most brokerage owners already suspect: the gap between what brokerages buy and what agents actually use is widening. Brokerages have rapidly added AI-branded products to their tech stacks. Agent usage is shallow and concentrated in just a few categories. The agents reporting significant productivity gains aren't relying on the full menu of brokerage-provided AI. They're using general-purpose chatbots and a few tightly integrated tools.
The pattern is clear enough that it has a name: the AI productivity gap.
What brokerages are buying
Brokerages are rolling out AI inside CRMs, marketing suites, listing-photo tools, and back-office systems, often marketed as end-to-end AI platforms. These typically include auto-writing listing descriptions, social posts, listing-photo edits, smart drip campaigns, and basic document summarization.
- AI-enhanced CRMs and lead-routing tools
- Marketing and social content generators
- Photo and video tools (virtual staging, image enhancement, video snippets)
- Document tools (summarizing contracts, extracting data, form-fill help)
Brokerage leaders often justify these purchases with broad promises of productivity and competitive edge. Few have a clear metric for agent-level adoption or income impact.
What agents actually use
Inman Intel data shows the majority of agents who use AI rely first on free or low-cost general chatbots (ChatGPT, etc.) to write text for listings and marketing, not on proprietary brokerage systems. The most common real-world uses are writing or improving property descriptions, drafting social media posts and email copy, quick market-data summaries, and simple image generation for social content.
Power users — the small group reporting significant productivity gains — go further and use AI for summarizing contracts, generating documents, automating client communication, and deeper market analysis. But many of these workflows still run through third-party tools or custom setups the agents chose themselves, not the default brokerage stack.
The takeaway is blunt: agents are not waiting for brokerage IT to solve their problem. They're solving it themselves with tools that cost nothing or $20 a month.
What actually produces ROI for a Pittsburgh brokerage
The patterns that correlate with real ROI are specific. Tools that live where agents already work (inside the CRM, email, or transaction system) rather than on yet another separate portal. Workflows that start from a concrete business problem: speed up contract review, prioritize leads, get three more listing appointments per month. Not use more AI features.
Systems that turn data into daily direction — surfacing who to call, what to send, what to say — instead of just generating more generic content. The tools that pay off are those that eliminate friction in a money-critical workflow (lead follow-up, listing prep, negotiation, client communication) and can be measured in hours saved or revenue influenced.
For a South Hills brokerage or team, that means starting with a usage audit. Identify which AI tools agents are actually opening each week, and cut back on the rest. Anchor AI to 2-3 revenue-critical workflows and design clear playbooks around those, not around the tool catalog. Invest more in training, coaching, and real workflow design than in adding new AI vendors. Power users emerge where support and experimentation are encouraged.
Define and track specific metrics like time-to-contract, response time to new leads, and follow-up completion rates so you can actually see ROI instead of just logins.
Common questions
If my agents are already using ChatGPT, why would they need an AI-integrated CRM?
They probably don't. If your agents are getting what they need from ChatGPT, the real value of an integrated tool is automation and consistency — not a new interface. The question to ask is: what workflow would save an agent 2-3 hours a week if it ran without daily intervention? That's where integration earns its cost.
How do I know if my brokerage is in the AI productivity gap?
Run a simple audit. Ask your agents which AI tools they actually open in a typical week. If the list is shorter than what you're paying for, you're in the gap. If they're using tools you didn't provide, they're solving problems your stack didn't address. That's the signal.
Should I stop buying AI tools for my team?
Not necessarily. Stop buying tools hoping they'll solve productivity problems on their own. Start buying tools that solve a specific workflow problem you've already identified and measured. The difference is huge.
If you're a Pittsburgh brokerage owner or team leader trying to figure out which AI workflows actually move the needle for your agents, let's talk. An audit of your current stack and a clear playbook for 2-3 high-impact workflows beats another software subscription every time.