A Bad Flight Costs Four Hours; a Bad Neighborhood Costs the Trip
WEEK 95 :: POST 3 :: CLAUDE
Directions Given To The A.I. This Week+
Instructions Given to each A.I. — Please provide 3 prompt variations that share this objective:
Each A.I. also received two static attachments: the blog post template (structure) and the authoring instructions (voice and standards). The text below is the week-specific assignment as sent — reflowed for the web; wording unchanged.
I'd like you to write this week's Ketelsen.ai post. Two files are attached: the blog post template (the structure to follow) and the authoring instructions (context, voice, and standards). Please read both before you begin, then produce the complete post in a single response.
This week's theme: "Where You Sleep Changes Everything" — Lodging and Neighborhood Intelligence.
This is Week 4 of an eight-week series on planning a vacation with AI. Week 1 established the reader's real constraints — a validated budget ceiling and a constraint profile. Week 2 chose the destination. Week 3 locked the flights, which fixed the dates and the arrival airport. This week they decide where they will actually wake up every morning.
Lodging is the decision travellers most often get wrong in a way they cannot undo. A flight that disappoints is four hours; a badly chosen neighbourhood is the whole trip. It is also the part of travel where the listing is written by someone whose interest is opposed to the reader's: the price shown is not the price paid, the photographs are chosen and cropped, the reviews are a mixture of the genuine, the incentivised, and the fabricated, and the words that sound most reassuring — cozy, vibrant, up-and-coming, steps from — are the ones most often doing concealment work.
The job this week is to give the reader a defensible lodging decision: this property, in this neighbourhood, at this true all-in cost, with these cancellation terms, chosen for reasons they can state.
The three prompts should help a reader work through:
- True total cost, not nightly rate. Resort fees, cleaning fees, service fees, occupancy taxes, parking, security deposits, mandatory "destination charges" — the drip-pricing pattern from Week 3's airfare work reappears here in a different costume, and it is usually worse. Two properties whose headline rates differ by 15% can invert once everything is added.
- Review forensics. How to read a review set rather than a review score: weighting recent reviews far more heavily than old ones, spotting clusters of suspiciously similar phrasing or timing, reading the three-star reviews where the honest detail lives, and noticing what a stream of positive reviews never mentions. A property whose complaints are all about the same thing is telling the reader something a 4.5 average conceals.
- Neighbourhood, decided by block and not by city. Walkability to what the reader actually plans to do, transit access at the hours they will use it, noise, and how the character of an area changes between a Tuesday afternoon and a Saturday night. The right question is never "is this a good area" but "is this a good area for this trip, these travellers, and these hours."
- Cancellation-policy risk as a priced feature. A non-refundable rate is a discount in exchange for accepting a risk. The reader should be able to say what that risk is worth to them rather than defaulting to the cheaper number.
- Decoding listing language. What cozy, charming, lively, convenient for transport, and partial view reliably mean, and which specific photograph or floor-plan absence should prompt a direct question to the host.
The output a reader should walk away with is a ranked shortlist they can act on: two or three properties scored on true cost, review credibility, location fit, and cancellation risk, with the reason each one ranks where it does.
A note on the strongest version of this week: at the advanced end, this is a lodging dossier matrix — candidate properties as rows; true all-in nightly cost, review-pattern credibility, location score against this trip's actual itinerary, and cancellation-policy risk as columns; with the reader's own weightings applied and the trade-offs made explicit. That structure is worth reaching for, and it is the natural successor to Week 3's fare decision framework.
A hard constraint, carried forward from Week 3 and just as binding here. AI models cannot see live listings, current rates, or today's availability, and their knowledge of specific properties is stale, thin, and — for anything below the level of a famous hotel — frequently invented. No prompt in this post may ask the AI to recommend a named property, quote a current rate, describe a specific listing it has not been shown, or assess a named hotel's present condition. A confidently hallucinated hotel recommendation is worse than a hallucinated airfare: the reader books it.
Neighbourhood character deserves particular care. A model will happily describe a district's safety or atmosphere from training data that is years stale and was never reliable — and this is the one place in the series where a wrong answer can put someone somewhere they should not be. Prompts should have the AI generate what to check and where to check it — the questions to ask, the hours to look at, the sources to consult — rather than deliver a verdict.
Design the prompts so the AI does what it is genuinely good at: structuring the comparison, exposing the fee layers, teaching the reader to read a review set, and naming what to verify. The reader supplies the listings; the AI supplies the judgement framework. Posts that blur that division should expect to be marked down on Practical Utility, exactly as in Week 3.
Series dependency chain, for the Metadata block: Week 4 consumes the confirmed routing and dates from Week 3, the destination from Week 2, and the budget ceiling from Week 1 — lodging is scored against what the flights left of the budget, and a nightly rate that breaks the ceiling is a signal to revisit the property tier or the neighbourhood, not to quietly raise the budget. Week 4 produces the booked lodging and its location, which Week 5's itinerary assumes as its starting point every morning: the itinerary is built outward from where the reader wakes up.
Because readers may arrive at this post without having read Weeks 1 to 3, the prompts should work for someone who knows their destination, dates, and rough remaining budget, while making clear they get far more from them with a real constraint profile, confirmed flights, and a live set of candidate listings in hand.
Three difficulty tiers as always — Beginner, Intermediate, Advanced — each a genuinely different approach to the same problem, not the same prompt at three lengths.
On examples: this is a consumer travel topic. The template lists tech startup / retail / freelance as suggested industry examples — those are marked MAY, and this week you should almost certainly adapt them. Families needing two bedrooms and a kitchen, couples choosing between a central hotel and a quieter rental, solo travellers weighing safety and walkability, older travellers for whom stairs and lift access decide everything, and anyone booking around a fixed-date event are the right contexts here. Choosing them over the suggested business examples is correct behaviour and will not be scored against you.
A note on supplied figures. Anything marked `[SUPPLIED — use as given]` above came from Ketelsen.ai's own research brief. Use it freely — you are not fabricating by repeating it, and you will not be marked down for leaving it uncited. Do not attach an invented source to it. (No supplied figures this week. Given the live-listing constraint above, this is a bad week to invent any — if you find yourself reaching for a typical resort fee or an average nightly rate, that is the signal to restructure the prompt so the reader supplies the real number instead.)
## BEFORE YOU SUBMIT — STRUCTURAL CHECK
(This block is identical every week. It exists because these specific items are the ones posts drop, and a dropped structural item costs compliance points for something that takes one minute to add.)
Your post is parsed by a script before any human reads it. Confirm all seven:
1. ☐ Response begins with `PLATFORM: <your name>` and `WEEK: 4` 2. ☐ `## Lead` present once, at the very top, before Variation 1 3. ☐ `## In one line` present in all three variations 4. ☐ `## What this prompt gives you` present in all three variations 5. ☐ `## The Prompt` present in all three variations, with the prompt in double quotes beneath it 6. ☐ `## Introductory Hook` and `## Current Use` present in all three variations (three of each — not one) 7. ☐ Every template heading written as `##`, none bolded instead; prompt breakdown is running text split on ` : `, with no `###` headings inside it
A complete post has 57 `##` headings. If your count is well short, a section is missing or was bolded instead of hashed.
One extra check this week: confirm no prompt asks the AI to name, rate, or price a specific property, or to pronounce on a neighbourhood's current safety. Those must be things the reader goes and verifies.
A flight that disappoints costs you four hours; a badly chosen place to stay costs you the whole trip — and unlike almost every other travel decision, it's the one written entirely by someone whose interest runs opposite to yours. This week's three prompts hand the judgment back to you: a Beginner prompt that strips a listing down to the true all-in price, an Intermediate one that teaches you to read a review set and build a neighborhood verification plan, and an Advanced one that turns two or three candidates into a scored, defensible decision. None of them ask an AI to pick your hotel — that would be the fastest route to somewhere you shouldn't be. They make you the traveller who can say exactly why this place, on this block, at this price.
The All-In Price Reveal
Turn advertised nightly rates into the total you'll actually pay.
The number on the listing is not the number on your card. Between them sits a quiet layer of cleaning fees, service fees, resort or "destination" charges, occupancy taxes, parking, and refundable deposits that never appear in the headline — the same drip-pricing trick you met in last week's airfare work, wearing a different costume and usually behaving worse. Two places whose advertised rates differ by fifteen percent can swap places once everything is added, so the cheaper-looking option quietly becomes the more expensive stay. This first prompt does the one thing that fixes it: it forces every option into a true all-in total, shows the arithmetic, and tells you which fees to go dig for before you commit.
Why this matters now
Regulators have finally noticed the problem, which tells you how bad it got. As of May 2025 the FTC requires hotels and short-term rentals to show an all-in total up front — but enforcement is uneven, older listings lag, and taxes and some charges can still land at checkout. That makes right now the exact moment this prompt earns its keep: you can no longer assume the headline rate is honest, and you can't yet assume it's complete. Running two or three options through a true-total comparison before you book is the difference between choosing on price and being surprised by it.
I'm comparing a few places to stay for an upcoming trip, and I only care about the total I'll actually pay — not the nightly rate they advertise. I'll give you the numbers I can see; you do the math and tell me what I'm missing.
Here are my options. For each one I'll list a label, the nightly rate, the number of nights, and any extra fees, taxes, or deposits shown — cleaning fee, service fee, resort or destination fee, occupancy or city tax, parking, credit-card surcharge, refundable deposit. If I can't find a number, I'll write 'not shown.'
[Option A: ...]
[Option B: ...]
[Option C: ...]
Do four things, in order. First, for each option, add everything up into a true all-in total and a true cost per night, and show your arithmetic so I can check it. Second, list every fee category that lodging commonly adds at checkout or on arrival that I might not have entered above — the ones I should go back and hunt for before I decide. Third, tell me which option is genuinely cheapest once everything is counted, by how much, and flag any option whose lower advertised rate becomes the more expensive stay. Fourth, ask me for the one or two missing numbers most likely to change the ranking. Do not guess or fill in any fee I didn't give you — where a number is missing, label it 'confirm before booking' rather than inventing an amount.
How the AI reads this prompt
Practical examples from different industries
A family of four choosing between two rentals for a week. The parents have two listings open: one at $180 a night that looks plain, one at $150 a night that photographs beautifully. They paste both into the prompt with every fee they can see. The AI's arithmetic shows the $150 place carries a $250 cleaning fee and a $95-per-night "amenity charge," pushing its true nightly cost to $186 — above the "expensive" option — and flags that neither listing showed occupancy tax, which the family then confirms adds roughly nine percent to both. They book the honest-looking one, having nearly chosen on a rate that was fiction.
A couple weighing a downtown hotel against a rental across the river. The hotel's rate looks steep until the prompt itemises what it includes, while the rental's lower headline hides a service fee, a cleaning fee, and paid parking they'd assumed was free. Running both through the true-total comparison, the couple discovers the gap between them is a third of what the advertised rates suggested — small enough that the decision now turns on location and cancellation terms rather than price, which is exactly where they wanted the decision to sit. The prompt didn't pick for them; it cleared away the false signal so the real trade-off could surface.
A solo traveller booking three nights around a friend's wedding. With a fixed date and no flexibility, price discipline matters more than usual. The traveller lists four candidate rooms with whatever fees are visible; the AI produces four true totals and, crucially, a note that two listings didn't display any tax at all — a warning to confirm before assuming they're the bargains they appear. Armed with the comparison and the "confirm before booking" flags, the traveller emails two hosts for the missing numbers and books the option that's genuinely cheapest once every line is counted, not the one that merely looked it.
Creative use case ideas
Reuse the same structure for a rental car quote, where "mandatory" airport surcharges, young-driver fees, and insurance add-ons play exactly the role resort fees play in lodging. It works for a gym membership comparison, too — join fees, monthly rates, and cancellation penalties hide the true cost of a year in the same way. Wedding vendors are a natural fit: caterers and venues quote a per-head number that balloons once service charges and rentals are added. For a non-travel, personal use, run it on two phone plans or two streaming bundles, entering the promotional rate and every add-on, to see which is actually cheaper after the introductory period ends. And for a community group booking a hall for an event, the prompt turns a confusing quote sheet into a single defensible number the treasurer can approve.
Adaptability tips
Scale it down by pasting a single listing and asking only "what fees am I likely not seeing here?" — a thirty-second gut-check before you fall for a rate. Scale it up by adding a fifth field for loyalty points or credit-card rebates, so the true total reflects what you'll effectively pay rather than the sticker. If you're travelling abroad, add "convert everything to [my home currency] at today's rate, and note that the conversion is approximate" so the comparison lands in money you actually think in. You can also change the objective line — swap "the total I'll actually pay" for "the total per person" when splitting a group rental, and the whole comparison reshapes around the number that matters to your group.
Pro tips
Ask the AI to sort its final answer from lowest true total to highest and to state the dollar gap between each — a ranked list is easier to act on than a paragraph. Add "flag any fee that is refundable versus non-refundable separately," because a $300 deposit you get back is not the same as a $300 fee you don't, and a naive total treats them alike. If you're comparing more than three options, ask for the results as a simple list with one line per property rather than prose, so you can scan it. Finally, keep the AI's fee-category checklist from step two — it's reusable, and pasting it into your next booking search means you already know what to look for before you even open the listing.
Prerequisites
Have the listings you're comparing open in front of you, and be ready to type out the numbers you can see: nightly rate, number of nights, and any fees, taxes, or deposits displayed. You don't need the complete picture — missing numbers are handled by the prompt — but the more real figures you supply, the sharper the comparison. No account, plugin, or paid tier is required; any general-purpose AI chat window will run this.
Required tools
Any general-purpose AI chat assistant on its free tier — ChatGPT, Claude, Gemini, or similar. No web browsing, plugins, or paid features needed, since you supply the numbers yourself.
Frequently asked questions
Doesn't the new FTC rule mean listings already show the total, making this unnecessary?
Not reliably yet. The rule took effect in May 2025 and does push platforms toward all-in pricing, but compliance is uneven across sites, some taxes and charges can still appear at checkout, and rules vary by country. Treating the headline as complete is exactly the assumption that gets travellers burned, so a quick true-total check is still worth the two minutes.
What if I genuinely can't find a fee the AI says to look for?
That's a useful result, not a failure. A fee you can't locate is a question for the host or platform, and the prompt is designed to surface exactly those gaps as "confirm before booking" items. Email the host with the specific question — "is there a cleaning fee or resort fee not shown in the listing?" — and you'll often flush out a number that changes your decision.
Can the AI just find these listings and fees for me?
No, and you shouldn't let it try. An AI can't see live rates or current availability, and if you ask it to supply the numbers it will invent plausible-looking ones. This prompt works precisely because you provide the real figures and the AI provides the structure and the gap-check — keep that division and the answer stays trustworthy.
The two options came out nearly identical. Now what?
Then you've learned something valuable: price isn't your deciding factor, so stop optimising it. When true totals land within a small margin, move the decision to the things price was hiding — location fit, cancellation terms, and review credibility, which the next two prompts are built to handle.
Recommended follow-up prompts
Run Variation 2, "The Review Detective," on whichever option wins on price, to make sure the cheapest true total isn't cheap for a reason the reviews would have told you. Then, if you're torn between two or three finalists, hand them to Variation 3, "The Lodging Dossier Matrix," which folds true cost together with reviews, location, and cancellation risk into a single scored decision. You can also pair this with the Ketelsen.ai airfare true-cost prompt from Week 3 to see your combined flights-plus-lodging spend against the budget ceiling before you commit to either.
Tags and categories
Tags:
lodging, hotels, short-term rentals, true cost, resort fees, drip pricing, budgeting, travel planning, comparison shopping, beginner prompts Categories: Travel Planning, Budgeting & Money
Citations
Federal Trade Commission, "Rule on Unfair or Deceptive Fees" (the "Junk Fees Rule"), 16 C.F.R. Part 464, finalized December 17, 2024, effective May 12, 2025 — requires hotels and short-term lodging to disclose an all-in total price and prohibits burying mandatory fees. FTC press release and small-entity compliance FAQ, ftc.gov.
The Review Detective
Read a review set like an investigator, then verify the block yourself.
A 4.5-star average is a summary designed to stop you reading, and that's the problem. Review scores flatten the one thing that would actually help you — the pattern in the complaints — into a single reassuring number. Meanwhile the listing's own words are chosen by someone selling to you: "cozy" and "up-and-coming" and "convenient for transport" are the phrases most often doing concealment work. This prompt refuses to accept the verdict and rebuilds the evidence instead. It teaches the AI to translate the listing's language, to read a review set the way a skeptic would — recent over old, honest three-star over glowing five-star, patterns over averages — and, critically, to hand you a checklist for verifying the neighborhood yourself rather than pronouncing on it from stale memory.
Why this matters now
The review ecosystem got measurably more polluted, and the law caught up. In October 2024 the FTC's rule banning fake and AI-generated reviews took effect, a direct response to how cheap it became to flood a listing with realistic fabrications. That cuts both ways for you: the worst fakes are now illegal, but they haven't vanished, and the ones that remain are better made than ever. Reading a review set by its patterns — clusters of similar phrasing, suspicious timing, the tell of what glowing reviews never say — is now a core traveller skill, not a paranoid one. This prompt turns that skill into a repeatable method you can run on any listing.
Act as a skeptical travel researcher who treats every listing as sales copy until proven otherwise. Your job is not to tell me whether this place is good — it's to help me separate what's verifiable from what I have to go check myself.
Here's what I can see about the one place I'm considering:
- The listing's own description, pasted verbatim: [paste]
- Its review score, the total number of reviews, and how recent the most recent ones are: [paste]
- A handful of actual review texts, including any three-star ones I could find: [paste]
- My trip: who's travelling, what we plan to do each day, and the hours we'll actually be coming and going: [describe]
Give me three things, clearly separated.
First, a translation of the description. For each reassuring phrase (things like 'cozy,' 'lively,' 'up-and-coming,' 'convenient for transport,' 'partial view'), tell me what it might be quietly working to conceal, and name the specific photo, measurement, or floor-plan detail I should ask the host for to confirm or kill the worry.
Second, a review-reading method applied to what I pasted. Weight recent reviews more heavily than old ones and say why; flag any cluster of suspiciously similar wording or timing; tell me what a wall of glowing reviews conspicuously never mentions; and pull the honest signal out of the three-star reviews, which is usually where it lives.
Third, a neighborhood verification checklist built for my trip. Give me the exact questions to answer, the specific hours of day to check, and the public sources or tools to use, so I can judge whether this block works for these travellers at these hours. Do not tell me whether the area is safe or describe its current character from memory — you don't have current, reliable information on that, so hand me the checks to run, not the verdict.
How the AI reads this prompt
Practical examples from different industries
An older couple who need step-free access. Their whole trip depends on stairs, and listings are notoriously coy about them — "charming townhouse" can mean four flights and no lift. They paste the description and reviews, and the prompt's translation pass flags "characterful" and "split-level" as phrases to interrogate, with a specific instruction to ask the host for a photo of every entrance and interior staircase. The review method surfaces a buried three-star note about "a lot of steps," which the glowing reviews never mention. They ask, they learn there are two flights to the bedroom, and they keep looking — a trip saved by reading the silence.
A solo traveller weighing walkability and evening comfort. She'll be out at night and back late, so "convenient for transport" and "vibrant area" are exactly the phrases she needs decoded. The AI translates "vibrant" as potentially "loud until 2am" and turns her concern into a checklist: check the walking route from the nearest transit stop at the hours she'll use it, look at street-level imagery for lighting and foot traffic, and consult the platform's own map rather than trusting the listing's "5 minutes from downtown." It never tells her the area is safe — it hands her the checks so she can decide with current information, not the model's stale guess.
A family renting around a specific festival with fixed dates. Their reviews skew heavily toward one season, and the prompt catches it: the AI notes that nearly all recent reviews cluster around summer stays and that nothing mentions what the neighborhood is like during the festival crowds they'll actually be there for. That absence becomes a verification task — check local sources for road closures and noise during the event, and ask the host directly how the block behaves that specific weekend. The family learns the street is on a parade route, information no star rating would ever have told them.
Creative use case ideas
The same detective method works on restaurant reviews before you book a special dinner — weight the recent ones, read the three-star reviews for honest detail about service and noise, and notice what the raves never mention. It transfers cleanly to choosing a contractor or a dentist, where a wall of five-star reviews and a few damning three-star ones tell very different stories. Use it on product reviews before a big purchase to spot suspiciously clustered praise. For a personal, non-commercial use, run it on reviews of a hiking trail or campground before a family trip, decoding "moderate difficulty" and building a checklist to verify conditions for the season and the ages in your group. Community organisers can even apply it to venue reviews before booking a hall, reading past complaints for the logistics that photos hide.
Adaptability tips
Tighten the focus by naming your single biggest worry up front — "my priority is quiet at night" — and the AI will weight its whole analysis toward noise-related signals in the reviews and the verification list. Broaden it by pasting reviews from two or three platforms at once and asking the model to note where they agree and where they diverge, since a property praised on one site and panned on another is telling you something. For international trips, add "assume I don't know local norms, and flag anything a local would find obvious," which surfaces the context you'd otherwise miss. And if you're short on reviews to paste, say so — ask the AI to tell you what the thin review count itself implies and what to verify because the reviews can't.
Pro tips
Paste the dates on each review if the platform shows them; timing clusters are far easier for the AI to flag when it can see them. Ask it to separate "concerns I can verify before booking" from "concerns I can only check on arrival," because the first list should change your decision and the second should shape your backup plan. When you get the neighborhood checklist, run the checks yourself the same day — street-level imagery and transit schedules change, and a checklist acted on a month later is worth less. If a host dodges a direct question you drew from the analysis, treat the dodge as data; the prompt's whole purpose is to generate questions whose answers, or non-answers, tell you something.
Prerequisites
You'll need the listing text and a genuine sample of reviews copied out — including, ideally, a few three-star ones, which take the most digging and carry the most signal. Have a clear sense of your trip's shape: who's travelling, what you'll do, and the hours you'll be moving around, since the verification checklist is built against those specifics. A little willingness to then go run the checks yourself is part of the deal — this prompt produces a plan, and the plan only pays off if you follow it.
Required tools
Any general-purpose AI chat assistant will run the analysis. To act on the neighborhood checklist you'll want a maps tool with street-level imagery and a transit schedule source — both freely available — but the AI itself needs no plugins or paid tier.
Frequently asked questions
Why won't the AI just tell me if the neighborhood is safe?
Because it can't do so honestly. A model's knowledge of any specific area is years out of date and was never reliable to begin with, and a confident wrong answer about safety is the one mistake in trip planning that can actually endanger you. The prompt deliberately trades the verdict for a checklist so you make that call on current information you gather yourself.
How many reviews do I need to paste for this to work?
More is better, but even five or six real ones — especially a mix of recent and critical — give the method something to work with. What matters more than volume is authenticity and range: a handful of genuine three-star reviews is worth more than fifty glowing ones, because the honest detail lives in the middle of the distribution, not at the top.
The AI flagged a "suspicious cluster." Does that mean the reviews are fake?
Not necessarily — it means the pattern is worth a second look. Similar wording can come from a template the platform suggests, or a burst of genuine stays after a busy weekend. Treat the flag as a prompt to read those reviews more carefully and weigh them less heavily, not as proof of fabrication.
Can I use this if the listing has almost no reviews?
Yes, and it's arguably more important then. Ask the AI what a thin or brand-new review history implies and what you therefore can't learn from reviews and must verify another way. A new listing isn't automatically bad, but it shifts more of the burden onto the description translation and the neighborhood checks.
Recommended follow-up prompts
Feed a property that passes this review scrutiny into Variation 3, "The Lodging Dossier Matrix," so its review credibility becomes one scored column alongside cost, location, and cancellation risk. Pair it with Variation 1, "The All-In Price Reveal," to confirm the well-reviewed option is also honestly priced — the two checks together catch most of the ways a listing misleads. You might also try a Ketelsen.ai host-message drafting prompt to turn the specific questions this analysis generates into a polite, effective note to the host.
Tags and categories
Tags:
reviews, review analysis, fake reviews, listing language, neighborhood research, walkability, verification, short-term rentals, intermediate prompts, travel safety Categories: Travel Planning, Research & Verification
Citations
Federal Trade Commission, "Trade Regulation Rule on the Use of Consumer Reviews and Testimonials," 16 C.F.R. Part 465, finalized August 14, 2024, effective October 21, 2024 — prohibits fake and AI-generated consumer reviews, review suppression, and undisclosed insider reviews. FTC press release and Federal Register notice, ftc.gov and federalregister.gov.
The Lodging Dossier Matrix
Score your finalists into one weighted, defensible lodging decision.
By the time you're down to two or three real candidates, the decision stops being about any single number and becomes about trade-offs — the cheaper place is farther out, the better-reviewed one has a brutal cancellation policy, the perfect location costs more than the flights left in the budget. Holding four competing dimensions in your head at once is where good travellers make bad choices, defaulting to whichever option feels right or whichever price is lowest. This advanced prompt builds the structure that a gut call can't: candidates as rows, four weighted criteria as columns, your own priorities applied, and every trade-off named out loud. It produces a decision you can defend — to your travelling companion, and to yourself at 2am when you're wondering if you chose well.
Why this matters now
This is the natural successor to last week's fare-decision framework, and it matters now because lodging is the decision you most often can't undo. A disappointing flight is over in hours; a badly chosen neighborhood or a non-refundable booking in the wrong spot shapes every day of the trip. As the rest of the series has fixed your budget, destination, and dates, the remaining variables have narrowed to exactly the four this matrix scores — which is what makes a structured comparison worth the setup at this stage rather than earlier. Run it when you have finalists in hand and a companion to convince, and the decision arrives already justified.
Act as a decision analyst helping me choose lodging with a weighted, defensible comparison instead of a gut call. I'll give you two or three candidates and the raw information I've gathered on each; you'll turn it into a scored decision I can explain to whoever I'm travelling with.
For each candidate I'll provide a label; the true all-in cost per night and total (or the raw numbers for you to compute); the cancellation terms; the review score plus a few representative review texts; and the location details relative to my plans — what I need to reach, how I'll get there (on foot, transit, or car), and at what hours.
My trip context: [who's travelling; the destination; the confirmed dates; what's left of my budget after flights; and any non-negotiables — step-free access, a kitchen, two bedrooms, quiet at night].
Work in this order.
1. Ask me to assign weights, as percentages summing to 100, across four criteria: true all-in cost, review credibility, location fit for this specific itinerary, and cancellation-policy risk. If I don't give weights, propose a default set based on my trip context, explain your reasoning, and wait for me to accept or adjust before you score.
2. Score each candidate from 1 to 5 on each criterion, using only what I gave you. For any cell where you don't have enough information to score honestly, write 'unscored — need [the specific thing]' instead of guessing a number.
3. Lay out the comparison as a plain-text grid: candidates as rows; the four criteria as columns; each cell showing the raw 1-5 and its weighted contribution; a weighted total per candidate.
4. Below the grid, write the trade-off in plain words: which candidate leads on the numbers, what picking it costs me (the dimension where it's weakest), and what change in my weightings would flip the result to a different candidate.
5. Finish with a verification punch-list: every cell you marked unscored, and — for the leading candidate — the single most important thing to confirm before I book.
One hard rule throughout: do not invent fees, review content, cancellation terms, distances, or any judgment about a neighborhood's safety or current condition. Every score must trace back to something I gave you. Anything that can't becomes a line on the punch-list, not a guess.
How the AI reads this prompt
Practical examples from different industries
A family choosing among three rentals for a two-week stay. Their non-negotiables are a kitchen and two bedrooms, and their budget after flights is tight, so they weight cost at 40 percent, location at 30, reviews at 20, and cancellation risk at 10. The matrix reveals that their cheapest option scores lowest on location — a long transit ride from everything they've planned — and that a modest shift of weight toward location would hand the win to the mid-priced candidate. Seeing the trade-off named, they decide the daily commute isn't worth the savings and choose the middle option with their eyes open.
A couple deciding between a flexible-but-pricey hotel and a cheaper non-refundable rental, with uncertain travel plans. Because their dates might change, they weight cancellation risk unusually high, at 35 percent. The AI scores the non-refundable rental a 1 on that criterion and, in the trade-off summary, spells out that its price advantage evaporates the moment plans shift. The couple books the flexible option, having watched a discount they were tempted by get correctly priced as a bet they didn't want to make.
A solo traveller attending a conference on fixed dates, comparing two well-located options. Location fit and review credibility matter most, so those carry the weight. The matrix marks one candidate's neighborhood-at-night score as "unscored — need: check the walk from the venue after evening sessions," turning the one thing the traveller hadn't verified into an explicit task rather than an assumption baked into a number. She runs the check, confirms the route, and books — with the single most important unknown resolved before payment rather than discovered on arrival.
Creative use case ideas
The weighted-matrix structure generalises to any multi-candidate decision with competing criteria, so it works just as well for choosing between job offers scored on pay, commute, growth, and culture. Apartment hunters can run it on rentals across price, location, condition, and lease terms. It fits choosing a car across cost, reliability, mileage, and features. For a personal, non-business use, families can score summer-camp options for a child on cost, activities, distance, and reviews, weighting by what that particular child needs — the same discipline that saves a vacation saves a decision that matters far more. Community groups can even use it to pick among venues or vendors, producing a scored rationale a committee can approve without relitigating from scratch.
Adaptability tips
Add a fifth criterion when your trip has one — accessibility, pet policy, or workspace quality — and adjust the weights to sum to 100 again; the framework doesn't care how many columns it has. Run it twice with different weightings to see how much your priorities actually move the answer, which is a fast way to discover whether you're genuinely torn or just anxious. If a companion disagrees with a choice, have them supply their own weights and re-run it — arguing about weights is far more productive than arguing about conclusions. And when you have only two candidates, keep the structure anyway; the value is in the named trade-off, which two options need as much as three.
Pro tips
Save the filled matrix; it's the record of why you chose, and it settles second-guessing better than memory ever will. Ask the AI to state its confidence in each score alongside the number, so a shaky 4 doesn't carry the same weight as a solid one in your reading. When the totals come out close, treat the "what would flip it" analysis as the real answer — a decision that's robust to small weight changes is one you can book without agonising, while a fragile one tells you to gather the missing information first. If you're deciding as a group, run everyone's weights and show all the matrices side by side; the pattern of where they agree is usually the actual consensus.
Prerequisites
This prompt rewards preparation: you'll want the outputs of the first two prompts in hand — true all-in costs from Variation 1 and review credibility from Variation 2 — plus the cancellation terms and honest location details for each finalist. Have your trip context ready, including what's left of your budget after flights and your genuine non-negotiables, since the weighting step leans on them. It works with two or three candidates; more than that and you should shortlist first, because the matrix is a deciding tool, not a browsing one.
Required tools
Any capable general-purpose AI assistant handles the scoring and the plain-text grid. A reasoning-focused model tier tends to produce cleaner weighted arithmetic and sharper trade-off analysis, but no plugins, browsing, or special features are required — you supply all the inputs.
Frequently asked questions
Isn't a weighted matrix overkill for booking a hotel?
For a one-night stay, yes — use Variation 1 and move on. This prompt earns its setup when the stakes are real and the trade-offs genuinely compete: a long trip, a tight post-flight budget, finalists that each win on a different dimension. The structure exists precisely for the decisions where a gut call tends to go wrong and where you'll want a defensible reason later.
What if the AI and I disagree with a score it assigned?
Override it — the scores are yours to set, and the AI's job is to structure and total them, not to have the final word. If a score feels off, tell the model what you'd assign and why, and ask it to recompute. Disagreeing with a specific cell is exactly the kind of engaged scrutiny the matrix is designed to invite.
How do I keep it from just inventing the numbers it's missing?
The prompt's structure does most of that work: it confines scoring to your inputs and routes every gap to an "unscored — need X" line. If you ever see a confident score for something you didn't provide, that's your cue to push back and ask what it was based on. Treat any un-sourced number as a red flag, not a convenience.
The two top candidates tied. What does that actually mean?
It usually means your four criteria don't separate them and a fifth factor is doing the real deciding — bring it in as another weighted column. A tie can also be genuine, in which case the "what would flip it" analysis tells you which small preference to trust. Either way, a tie is information: it says stop optimising and pick on the thing the matrix didn't capture.
Recommended follow-up prompts
Once the matrix names your winner, run a Ketelsen.ai host-outreach prompt to send the pre-booking questions from your verification punch-list before you pay. Loop back to Week 3's airfare decision framework to confirm your combined flights-plus-lodging total still sits under the Week 1 budget ceiling — the matrix should never quietly push you past it. And keep your filled dossier for Week 5, where the itinerary gets built outward from exactly the location this decision just locked in.
Tags and categories
Tags:
decision matrix, weighted scoring, lodging, trade-offs, cancellation risk, location fit, advanced prompts, decision analysis, travel planning, structured output Categories: Travel Planning, Decision Frameworks
Citations
NOT APPLICABLE
Which of the three should you use?
The three prompts are a progression in ambition, not in length, and each is the right tool at a different moment in the lodging decision. The Beginner prompt, The All-In Price Reveal, does one thing completely: it converts advertised rates into true totals so you stop comparing fictions. Reach for it first, or on its own when price is your only real question and a two-minute check is all you want. The Intermediate prompt, The Review Detective, goes wider and deeper — it decodes the listing's language, teaches you to read a review set by its patterns rather than its average, and hands you a checklist to verify the neighborhood yourself. It's the one to run when a place looks good on paper and you want to know what the paper isn't saying.
The Advanced prompt, The Lodging Dossier Matrix, is different in kind rather than degree. Where the first two examine one option at a time, the matrix decides between finalists, folding cost, reviews, location, and cancellation risk into a single weighted comparison with the trade-offs named out loud. It expects the outputs of the other two as inputs, which is why it sits last: it's a deciding tool, not a browsing one, and it rewards the reader who arrives with real numbers already in hand.
They overlap deliberately at the seams — the matrix consumes the true costs from Variation 1 and the review credibility from Variation 2 — so running all three in sequence isn't repetition, it's assembly. A reader in a hurry can stop after the first and still book smarter than they would have. A reader facing a genuine, high-stakes choice between two neighborhoods and a fixed budget should run all three, in order, and end with a decision they can defend. Every one of them holds the same line: the reader supplies the listings, and the AI supplies the judgment framework — never the verdict.
TAGS:
