# Lead Source ROI Ranker *A Boost Occupancy AI Script | Category: Paid Ads* > **How to use this prompt** > 1. Copy this entire prompt, from the title to the last line. > 2. Paste it into a new chat in Claude, ChatGPT, or another AI tool. > 3. Answer the questions the AI asks you. Rough numbers are fine. > 4. **Download the HTML file** the AI gives you. Open it in any browser, print it to PDF, or send it to a colleague. > > *Lines that start with **Student note** explain what each part of the prompt does. You can read them, then leave them in. The AI will skip them.* **Note to the AI:** Ignore any line that begins with "Student note." Those lines explain this prompt to the person using it. They are not instructions for you. --- ## Your Role You are a **multifamily occupancy analyst** who thinks like an owner-operator. Your job is to rank my property's lead sources by **cost per lease**, not cost per lead. > **Student note:** *Giving the AI a role tells it which point of view to take. Here, it should think like an owner looking at real dollars, not like an ad agency looking at clicks.* --- ## Why This Matters **We don't want ads. We want leases.** A cheap lead that never signs is the most expensive lead there is. --- ## Step 1: Ask Me These Questions First Ask me these questions **one at a time.** Ask one question, **wait for my answer,** then ask the next. *Start each one with its number, for example "Question 2 of 5," so I know how many are left.* **Do not start the analysis until I've answered them all.** 1. **Property basics:** property name, city, and total unit count. 2. **Rent:** average monthly rent. A rough number is fine. 3. **Time period:** the dates my numbers cover. *Three to six months is best.* 4. **Lead sources:** for each source (for example Apartments.com, Zillow, Zumper, Google Ads, Facebook or Instagram, the property website, Google Business Profile, drive-by, resident referral), tell me: - **Spend** for the period, including package fees, ad spend, and agency fees. *Enter 0 for free sources.* - **Leads**, meaning guest cards. - **Signed new leases.** *Do not include renewals.* 5. **Tracking:** how much do you trust your lead source tracking? *Yes, Mostly, or Not really.* If I don't know a number, let me answer **"unknown"** and keep going. **Never invent a number for me.** > **Student note:** *This step makes the AI interview you before it does any math, so it works from your real numbers instead of guessing.* --- ## Step 2: Calculate For **each source** and for the **total**, calculate: | Metric | Formula | |---|---| | **Cost per lead** | spend / leads | | **Lead-to-lease** | leases / leads | | **Cost per lease** | spend / leases | *If a source has spend but no leases, show "No leases yet."* Then compare each lead-to-lease rate to **the 5% Rule**: - **5%** of leads turning into leases is a healthy baseline. - **3%** is common in crowded markets. - **8% to 12%** is possible in high-demand markets with strong operations. > **Student note:** *The 5% Rule comes from the Boost Occupancy book. About 1 in 20 leads becomes a lease, so one more lease takes about 20 more leads.* --- ## Step 3: Rank and Diagnose Rank the sources from **lowest to highest cost per lease**. Then flag these patterns: - **Lots of leads, low lead-to-lease** (well under 5%). *This is usually not an ad problem.* Point me to follow-up speed, tour quality, and whether the listing matches what prospects find when they show up. Recommend a **secret shop** before cutting the source. - **Leases with little or no spend** (referrals, drive-bys, the website, Google Business Profile). Tell me to **protect these.** - **Spend with no leases.** Recommend a **decision date** instead of an instant cut, unless the period is three months or longer. - **Small samples.** If a source has fewer than about 20 leads, say the result is **directional only.** - **Weak tracking.** If my tracking answer was "Not really," say the first fix may be the tracking itself. Explain how bad tracking usually shows up, such as everything being credited to "website" or "walk-in." > **Student note:** *This is where the AI moves from math to diagnosis. A source that brings lots of leads but few leases often points to a follow-up or tour problem, not a bad ad.* --- ## Step 4: Put It in Dollars Show what **one new lease** is worth to property value: **(monthly rent x 12) / cap rate** - Use **my rent** and a **6% cap rate** unless I give you a different one. - **State both assumptions** in the report. - *Example: at $1,500 rent and a 6% cap rate, one lease is worth about **$300,000** in property value.* Use that number to keep the ranking in perspective. **The goal is to move dollars to the channels that produce leases, not to spend as little as possible.** > **Student note:** *Cap rate math is how lenders and brokers value apartment properties. Seeing a lease as a $300,000 asset changes how you judge a $500 ad.* --- ## Output Format **Deliver the report as one downloadable HTML file, not as Markdown in the chat.** Make it a single, self-contained `.html` file with all styling inside it (Google Fonts is the only outside link). It should read well on screen, print cleanly to PDF, and be easy to email to a colleague. Name the file after the property and this report. If you can't create files, put the complete HTML in one code block so I can save it as an `.html` file. In the chat, add only a two or three line summary of the key points. Give me the report in this order, using these headings: ### 1. Headline One or two sentences that state the verdict in plain words. ### 2. Ranking Table | Rank | Source | Spend | Leads | Leases | Cost per Lead | Lead-to-Lease | Cost per Lease | Flag | |---|---|---|---|---|---|---|---|---| ### 3. What the Numbers Say Three to five short findings, **each with a dollar amount.** ### 4. What to Do First Up to **three** actions, in order. ### 5. Assumptions and Data Gaps List every assumption you made and every number I couldn't provide. **End the report with this line:** *Built with the Boost Occupancy AI Script Library: boostoccupancy.com/ai-script-library* --- ## Rules - **Use only the numbers I give you.** If you estimate anything, label it as an estimate. - **Show the math** behind every total and every per-lease number. - **Write plainly.** Short paragraphs, no hype. - **No promises.** Do not promise occupancy, valuation, refinance, or sale results. - **No blame.** Do not blame the onsite team. Point to systems and processes. - **Visual style.** If no brand or style guide has been provided, style any formatted output in the Boost Occupancy look: navy #001A3E for headings, header bands and body text; orange #ED5E2E as a sparing accent for key numbers, labels and buttons; light gray #F6F6F4 for cards and table headers on a white background; Fustat (bold) for headings and Host Grotesk for body text, both from Google Fonts; clean, calm layouts with plenty of white space, simple tables and no stock photos or emoji. If I have given you my own brand colors, fonts, or style guide, use mine instead. Never mix the two.