AI Connector (MCP): Workflows, Prompts, & Use Cases
Overview
The TractIQ AI Connector puts the industry's most complete self-storage dataset directly inside your AI conversation, Claude, ChatGPT, Copilot, and more. If you haven't set up the Connector yet, start here: AI Connector (MCP): Setup & Getting Started.
Whether you're evaluating an acquisition, benchmarking a portfolio, or building an Investment Committee memo, these workflows are designed to get you moving fast and show you what's possible.
The best way to start? Just ask. Open a new chat with TractIQ connected and type: "What can I do with the TractIQ Connector?" Your AI will walk you through everything available. The prompts below are starting points. Experiment and make them your own.
Quick tip: Specify "use TractIQ to..." in your prompts. Otherwise your AI may default to an internet search instead of pulling from the TractIQ Connector.
Instant Market Snapshot
Get a full read on any submarket in seconds. The Connector pulls supply metrics, demographics, and housing activity around any address.
Try prompting:
- "Using the TractIQ MCP, give me a market summary for [address] in a 3-mile radius."
- "Using the TractIQ MCP, what's the sq ft per capita, population, and median household income in the 5-mile trade area around [address]?"
- "Are there any facilities under construction within 5 miles of [address]?"
Competitive Analysis and Comp Sets
Build a full competitive picture for any location, current rates, unit mix, operator info, and rate trends.
Try prompting:
- "Find all self-storage facilities within 3 miles of [address] and summarize their current street rates by unit size."
- "Which operators are in my 5-mile trade area and what are they charging for a 10x10?"
- "Build a comp rate table: one row per competitor, columns for each unit size, street rate and web rate."
The Connector searches nearby facilities, pulls current pricing by unit size, and layers in historical rate trend data. Ask it to drop this into a spreadsheet and it will.
Deal Analysis from Your Saved Deals
If you've saved a deal in TractIQ, the Connector can find it and run a full analysis without you supplying coordinates or any additional context.
Try prompting:
- "Pull up my deal at [address or deal name] and give me a full market and competitive analysis."
- "What deals do I have saved in my [list name] list that are new or in underwriting stages?"
Building Full Underwriting Packages
This is where it really opens up. Investment memos, offering memorandums, pro formas, and assumption sets can all be assembled inside a single AI conversation, with TractIQ data populating the inputs automatically.
One important recommendation here: don't try to build a full underwriting package in a single prompt. Claude and ChatGPT work much better when you build up to it. Start by pulling comps, then run the rate analysis, then look at supply pressure, each as its own ask. Once you have those pieces, you can layer them into a bigger model. The results will be sharper and you'll stay in control of the output. You can also export large format datasets for demographics, housing/commercial developments, rental comps, etc from the TractIQ platform directly and then upload to your AI chat as extra conversational context.
Try prompting:
Underwriting models and pro formas:
- "Pull all storage comps within 3 miles of [address] and summarize current street rates by unit size."
- "Now run a rate trend analysis for those same facilities over the last 12 months."
- "Using that comp and rate data, build a Year 1 revenue assumption set for a [X] unit facility at [address]."
- "Build a 5-year pro forma using those assumptions, and flag any supply pressure that could affect the stabilization timeline."
Assumption sets:
- "Pull the demographic inputs I need for bad debt, concession, and discount assumptions for [market]."
- "Build an assumptions page, market rents by unit size, demographic summary, supply pipeline, and housing start trend."
Offering memorandums:
- "Draft a trade area brief for an OM, market overview, competitive landscape, demographic profile, and supply summary for [address]."
- "Generate an executive summary with key market stats, competitive positioning, and a supply pipeline callout for [address]."
The output lands directly in whichever tool you're working in, a Word doc, a PowerPoint deck, or a spreadsheet..
Pro tip: Upload your existing underwriting template and prompt: "Populate every field you can using TractIQ data for [address]." The AI will match TractIQ data points to your template's inputs and flag anything it needs you to fill in manually.
Working in Excel or PowerPoint? Use the Connector There Too.
If you're running Claude or ChatGPT as a plugin inside Excel or PowerPoint, the TractIQ Connector works natively in that context. TractIQ data gets pulled, structured, and written directly into your spreadsheet or deck in real time.
In Excel, ask it to build a new tab with comp rates, demographics, or market metrics and it will create and populate the sheet on the spot. Feed it your existing model and ask it to fill in TractIQ-sourced inputs directly into the relevant cells. Run multiple sheets in parallel from a single conversation.
In PowerPoint, ask it to build a market slide with a comp table, demographic summary, and supply narrative, assembled inline, inside the deck you're already working in.
For any workflow that lives in a model or a deck, running your AI as an Excel or PowerPoint plugin with the TractIQ Connector active is the fastest path. The data goes straight into the file.
CMBS Occupancy and Financial Performance
Available on Pro+ and MAX plans only. If you're on a Pro plan and want to explore an upgrade, reach out to us at support@tractiq.com and we'll walk you through the options.
For facilities with CMBS financing, the Connector surfaces occupancy history and financial performance data sourced through TractIQ's partnership with CRED iQ, covering 4,000 stores representing $50B in financed assets.
Try prompting:
- "Does [facility name] have occupancy data available? If so, show me the occupancy history over time."
- "Pull the NOI, revenue, and operating expenses for [facility name] and summarize the financial performance."
- "Find all facilities in my market or MSA that have CMBS performance data and compare their occupancy trends."
Portfolio and Pipeline Analysis
The Connector can work across your entire saved deal set at once, useful for portfolio reviews, pipeline check-ins, or flagging risk across multiple markets.
Try prompting:
- "Compare supply pressure and sq ft per capita across all the deals in my [list name] pipeline."
- "Which of my active deals are in the most undersupplied markets?"
- "Flag any deals in my pipeline where there's a facility under construction within 3 miles."
- "Build a summary sheet comparing demographics and competitive density for every deal I have in Underwriting."
Prompting Tips That Actually Make a Difference
You don't need to be precise or technical to get good results. That said, a few habits will consistently get you to better output faster.
Build up to complex asks, don't start with them. For anything multi-part, a full underwriting model, a portfolio comparison, a complete OM section: start with the first piece you need, review it, then ask for the next. The AI works best as a back-and-forth, not a one-shot brief.
Attach a reference example every time. A PDF or Word doc of a prior memo, deck, or model is worth more than three paragraphs of instructions. The AI reads it and matches the voice, structure, and visual rhythm automatically. It's the single highest-leverage move you can make.
Treat the first draft as the outline. Don't expect to ship the first version. Read it once, mark the 2-3 biggest issues, and send those back in one message. Two or three rounds of feedback usually gets you to something sendable.
Send one fix per message. Better to send one piece of feedback, see the result, then send the next. Piling multiple edits into one message dilutes the response.
Specificity beats adjectives. "Make it punchier" is weak. "Cut sections 1 and 2 entirely, open with the bold hook, close with the quote about pricing visibility" gets a great revision. Tell the AI what's wrong, not how to fix it — it'll propose the fix.
Ask the AI to grade itself. Before sending a deliverable, ask: "Will this land with [contact name]? Give me odds out of 10, honest." Claude in particular is well-calibrated and will surface things you should fix before it goes out.
Strip AI-tells on the final pass. Before anything goes out, ask the AI to remove words like "delve," "leverage," "pivotal," "robust," "navigate the landscape," and "in today's landscape." They scream AI to a sophisticated reader.
A worked example. Here's the actual opening prompt for a benchmarking memo that converted an institutional prospect:
"Build a trade-area competitive benchmarking memo for [client]. Their development site is at [address]. [Contact name] is the senior market analyst. [Description of the firm's strategy]. Attached is a prior memo for reference on format."
That's it. No section list, no chart specs, no word count, no colors. The reference memo gave the AI everything else it needed. Three rounds of short feedback got it to the final result.
Questions or ideas on how you're using it? We'd love to hear from you. Reach out anytime at support@tractiq.com.