Learn how to use AI well
The prompts are the starting point. This is what you need to know to get more out of them.
New: AI Connector operator guide
A 20-page PDF that shows how to turn your Storeganise data into analysis, actions, and ready-to-send writing with Claude or ChatGPT.
The AI progression
Most operators start with one-off prompts. That is a fine place to start. But there is a progression — and understanding it changes how you think about AI in your business.
Prompts
What it is: You type a request. You get an answer. You start again next time.
Skills
What it is: You tell the AI how to behave once. It remembers across conversations.
Projects
What it is: A persistent workspace loaded with your business context — your tone, policies, data.
Connectors
What it is: AI reads live data from your systems. Your FMS, your rates, your occupancy.
Automations
What it is: Multi-step workflows that run without you.
The platforms use different names for the same ideas. Do not let that confuse you. Where you start matters less than whether you keep building.
"Most operators are at Level 1 or 2. Competitive advantage starts at Level 3."
Prompt anatomy
Mediocre AI output usually means a mediocre prompt. Here is what separates a prompt that works from one that does not.
Role
Tells the AI who to be.
Context
Tells the AI what it needs to know.
Task
The actual ask.
Format
How you want the output.
Constraint
What to avoid or include.
"Write a late payment email"
Generic. Could be any industry. Unusable without heavy editing.
"You are an experienced self-storage manager at a 3-site portfolio. Write a late payment notice for a tenant 14 days overdue on a $99/month unit. Professional email, under 150 words. Firm but not aggressive. Do not mention legal action."
Specific. On-brand. Ready to send.
The element operators skip most often is Context. Add your facility name, your city, and the specific situation — watch the output transform.
Data security and safe use
AI tools are useful. They also handle information. Here is what you need to know before you put anything real into them.
Free plans and your data
Most free AI accounts use your conversations to train the model. Anything you type may be reviewed or used. For general questions and drafting, this is fine. For real tenant data, financials, or sensitive business information — use a paid account and check that the training data opt-out is on. It usually is not on by default.
Hallucinations — AI gets things wrong confidently
AI does not fact-check itself. It produces well-written, confident output that can be completely wrong. This happens most with specific numbers, dates, and factual claims. Always read before you use. The output is a starting point, not a finished product.
Connecting AI to your systems
If you connect an AI tool to your FMS using an API key, that connection gives the AI access to your data. Before you set one up, check what the connection can see — and whether it can write back to your system or only read from it. Read-only is the safer starting point.
A quick checklist
- • Are you on a paid account with training data turned off?
- • Does your prompt contain real tenant names, unit numbers, or financials? If yes, use placeholders while drafting.
- • If you are using a connector, do you know what data it can access?
None of this should stop you. It should make you thoughtful.
New to AI? Start here
You do not need a specific tool to use this library. Any of the main platforms will work. If you have not picked one yet, here is a simple way to think about it.
Claude
Strong for careful writing, long documents, and following detailed instructions. A good starting point for operators who want consistent, nuanced output.
Perplexity
Built for research and finding current information. Good for competitive research, market questions, and anything where you need sourced answers quickly.
ChatGPT
The most widely used platform. Large ecosystem of add-ons and integrations. A reasonable default if you are not sure where to start.
Free plans are a starting point, not a long-term solution.
Free plans limit usage and most use your data for training by default. Once AI becomes part of how you work, a paid plan is worth it.
Starting somewhere matters more than starting with the right one.
All three platforms are capable. The difference between them is less important than the consistency of use. Pick one and build on it.
What if you want to switch later?
You can. Your prompts work across platforms — copy and paste. The only thing that does not transfer automatically is your saved context. Ask the AI you are leaving to summarise everything it knows about your business into a document. Load that into your new platform as a starting point.
Building a competitive advantage
Every operator has access to the same AI tools. The difference is what you build with them.
Your prompts and context are an asset
A well-built set of prompts, projects, and instructions represents real operational knowledge. It reflects how your business works — your tone, your policies, your standards. That does not exist for your competitors. They start from scratch every time.
Consistency compounds
An operator using AI inconsistently gets occasional useful outputs. An operator who has built skills and projects gets consistent, on-brand output every time — without re-explaining context. That gap grows over time.
Smaller operators can now do what only large platforms could afford
Pricing analysis. Automated follow-up. Consistent marketing. Delinquency management. These used to require headcount. With the right setup, one operator with the right AI infrastructure can run a leaner, faster, more consistent operation than a competitor with twice the staff.
"The operators who treat AI as infrastructure — not a shortcut — will own the next decade."
The prompt library is where you start. The progression is how you build.