Introduction — Eleven Days

A faint smell on day seven, a stain on day eleven, and an insurer asking when it started.

---

The Booking

⚠ It was, by any measure, a good booking.

Eleven days. A three-bedroom house on a quiet road. Two cats — one friendly, one that lived under a bed and required nothing but a full bowl and to be left alone. A lot of plants. The post. The bins on Tuesday. An alarm with a four-digit code written on a card by the door.

⚠ The family had used a sitter before and had a system. There was a folder on the kitchen table with the vet's number, the neighbour's number, the boiler's service record, and a note about which window did not lock properly.

⚠ Sacha did it well. The cats ate. The plants lived. Nothing was late. On the last morning Sacha stripped the bed, ran the sheets, wiped the surfaces, and left the house tidier than it had been on arrival.

And on day seven there was a faint smell in the upstairs bathroom.

⚠ Damp, slightly. The kind of smell you notice and stop noticing within a minute.

Sacha thought about it for perhaps four seconds and concluded it was the towels. There was a bath mat that was never quite dry. It was raining that week. It was an old house.

⚠ None of that was unreasonable, and it was not written down anywhere.

---

The Stain

⚠ The family got home on a Sunday evening and found a brown stain spreading across the kitchen ceiling, about the size of a dinner plate.

⚠ Under the bathroom floor, a joint on a waste pipe had been weeping. Slowly. For some time — weeks, possibly months. It was nobody's fault. It was the kind of thing that happens in old houses and it is exactly what building insurance exists for.

Sacha was not accused of anything. The owners were decent people, they liked Sacha, and their first instinct was to apologise for the mess.

⚠ The insurer's questions were different, and they were entirely reasonable.

⚠ When did this start? Was anybody in the property? What did they observe? What was done about it, and when?

Those are not hostile questions. They are the ordinary questions any insurer asks, and they exist because the answer determines whether damage was sudden or gradual, whether it was mitigated, and whether the policy responds.

⚠ And Sacha could not answer a single one of them.

There was no daily log. No photographs. No note of anything at all — for eleven days in which Sacha had been the only person in that house.

⚠ Not a note of the smell on day seven. Not a photograph of the bathroom on any day. Not a record that the kitchen ceiling had been clean on day ten.

⚠ Sacha could say, honestly, that the ceiling had looked fine. Sacha could not show it.

---

What Actually Went Wrong

⚠ It is worth being precise, because the obvious lesson is the wrong one.

The obvious lesson is: report the smell. True, and small. Sacha should have mentioned it, and if Sacha had, the whole thing would have been a different story — a sitter who noticed something, said so, and prompted an early repair.

⚠ But Sacha did not fail to notice. Sacha noticed and dismissed it, which is what everybody does with a faint smell in an old house, and always will.

The real failure is structural.

⚠ For eleven days, Sacha was the only person who knew anything about the state of that house. Not the main source of information — the only source. And Sacha kept none of it.

⚠ Sacha had not failed to prevent a leak. Sacha had failed to be able to say when it started — and in an empty house, that is the same thing.

⚠ Because when the only witness has no record, the period is simply blank. Nobody can say the ceiling was clean on day ten. Nobody can say the bathroom smelled on day seven. Nobody can say whether anybody was in the house at all.

And a blank period is not neutral.

⚠ It is the period in which anything might have happened, and the only person who was there cannot say otherwise. That is a bad place to stand even when — especially when — you have done nothing wrong.

---

Why This Trade Is Different

⚠ Every trade in this series involves being trusted with something. This one involves being trusted with everything, alone, for a fortnight.

Consider what actually happens at a handover.

⚠ Somebody hands you keys to their home, the code to their alarm, and access to every room, every drawer, every document, every possession, and — frequently — a living animal that depends on you. Then they get in a taxi and go somewhere with a different time zone.

⚠ There is no supervision. There is no visit. There is no colleague. For the entire duration of the booking, the only account of what happened in that house is yours.

Compare that with anything else.

A cleaner is there for two hours, usually while somebody is in, and leaves a visible result.

An errand-runner is at a door for four minutes.

⚠ A house sitter has total unsupervised access, for days, to a property and its contents, with the owner absent — and everything that happens in that period is attributable to them by default, because there is nobody else it could be attributable to.

That is the whole business. Not looking after houses. Being the only account of a period in which nobody was watching, and being able to produce that account when somebody asks.

⚠ Which means the record is not administration. It is the product, it is your defence, and it is the reason the owner slept on the plane.

---

The Four Failure Modes

⚠ In this trade the serious failures come in four shapes, and only one of them is about doing the job badly.

One — something happened and nobody can say when. The leak. The break-in on an unknown night. The plant that died in week two. The pet that stopped eating on a day nobody recorded. Chapter 4.

⚠ Two — you did something in somebody's house that you were not entitled to do. Had a guest. Used the car. Slept in the wrong room. Opened a drawer. Worked from the study. Almost all of it innocent, almost all of it undiscussed, and all of it visible in hindsight. Chapter 5.

⚠ Three — access failed. A key lost. A code shared. A door left unlocked. Somebody else in the house who should not have been. An alarm not set. This is the failure that costs the most and the one people are most casual about. Chapter 6.

⚠ Four — an emergency, and you made a decision that was not yours. Authorised a repair. Made a call about an animal. Let somebody in. Decided it could wait. Chapter 7.

And underneath all four, the same thing. You were the only person there, and either you can account for it or you cannot.

---

The Four Rules

⚠ Rule 1: You are the only person who was there.

Everything that happens in that house while you hold the keys is attributable to you unless the record says otherwise. Which makes a dated daily log with evidence the single most important thing you produce.

⚠ Chapter 4, then Chapter 24.

⚠ Rule 2: The house is not yours.

You are in it, not of it. Guests, possessions, rooms, the car, the wine, the study, the bath. Every one of those has an answer and the answer is not "it was obviously fine."

⚠ Chapter 5, then Chapter 23.

⚠ Rule 3: Access is the whole job.

Keys, codes, alarms, and who else can get in. The single most valuable thing you are handed and the one most casually treated. Chapter 6, then Chapter 18 in full.

⚠ Rule 4: In an emergency you prevent harm and nothing else.

You contain, you call, and you record. You do not authorise, decide, repair, or judge. Chapter 7, then Chapter 25.

⚠ Everything in this book is one of those four, applied to a specific situation.

---

What Success Looks Like

⚠ Most people in this trade measure whether the booking went smoothly. That is a poor measure, because most bookings do, and because the ones that do not are frequently not your fault.

⚠ Every day you held the keys is accounted for, with dated evidence, so that any question about what happened in that house has an answer.

⚠ Two numbers.

⚠ Days with a complete log: target one hundred per cent. Not most days. Every day, including the ones where nothing happened — because the days where nothing happened are exactly the ones you will need to prove.

⚠ Access events by anybody other than you: target zero, and every single one recorded with who, when, why, and whether the owner authorised it.

⚠ Note what is not on that list. Whether anything went wrong is not a measure. Boilers fail, pets get ill, storms happen, and an old house does what old houses do. The measure is whether you can say what happened and when.

Chapter 24 builds the log. Chapter 32 makes it survive a busy week. Resource 29 is the sheet.

---

Where AI Fits, and Where It Does Not

⚠ This is an AI-powered business book, and the honest account of what that means here is narrower than the marketing.

⚠ AI is genuinely useful for the writing and the structure.

Building the property file questions — which are far more numerous than anybody expects and which owners never think to answer unprompted. Turning a rambling handover conversation into a written file. Drafting the agreement, the access clause, the emergency authorisation. Designing a daily checklist from a property's particulars. Writing update templates. Rehearsing the refusals. Modelling a booking calendar and a cost floor.

⚠ That is real value, because the paperwork is exactly what sitters skip and exactly what would have saved Sacha eleven days of unrecorded time.

⚠ AI is not a source of truth about property, animals, or emergencies.

⚠ It does not know whether a noise is a problem, whether an animal is unwell, whether a smell matters, whether a repair is urgent, or what your insurance covers. It will answer all of those fluently and it may be badly wrong, and the consequence lands in somebody's house or on somebody's pet.

⚠ Never ask it a veterinary question. Never ask it whether something is safe to leave until morning. Chapter 25 and Chapter 37.

⚠ And never put a property's address, alarm code, key location, or absence dates into it. That combination is a burglary brief, and it is the most dangerous data in this entire series of books.

---

How the Book Is Built

⚠ Thirty-seven chapters in five parts.

Part One — Getting Started (Chapters 1–7). The trade, the models, what this work is not, and one chapter per rule.

Part Two — Setting Up (Chapters 8–16). Refusals, credibility, the menu, niche, research, the plan, the money, the name, and the legal and insurance footing.

Part Three — Doing the Work (Chapters 17–26). The property file, access, the handover, the daily routine, pets, property routines, overnight stays, the log, emergencies, and departure. The operational heart.

Part Four — Pricing and Operations (Chapters 27–32). Packages, prices, agreements, the calendar, communication, procedures.

Part Five — Clients, Money, and Growth (Chapters 33–37). First clients, marketing, claims, money, responsible AI use.

⚠ Every chapter ends the same way: How AI Helps, Common Mistakes, Safety and Legal Considerations, Do This Now, and a Chapter Summary. Four prompts per chapter, one hundred and forty-eight in all, repeated in the library at the end.

⚠ Sixty-eight resources across six parts, referenced by number throughout and every one cited in a Do This Now list somewhere.

---

What This Book Will Not Give You

⚠ It names no platform, agency, insurer, alarm company, lock, camera, or product.

⚠ It states no law, licensing requirement, insurance position, tenancy or occupation rule, animal welfare requirement, or tax obligation.

⚠ This matters more here than in most trades. Whether staying in somebody's property creates any occupation right, whose insurance covers what while an owner is away, what obligations you take on by accepting responsibility for an animal, and what you may do in an emergency are all questions with real answers that differ by country and change.

⚠ Every one comes from your insurer, the owner's insurer, your accountant, a solicitor, a vet, or your local authority — obtained by you, dated by you, and written into Resource 3.

⚠ It gives no veterinary advice and no advice about buildings, systems, or repairs.

⚠ And every figure is in U.S. dollars, and every figure is blank.

---

A Note on the Warning Markers

⚠ The ⚠ marks are load-bearing.

⚠ They mark places where getting it wrong costs somebody their home's security, an animal's welfare, real property damage, real privacy, or your ability to keep trading.

⚠ They are not emphasis. Where you see one, slow down.

---

Before You Start

⚠ If you have already done a booking, do one thing before Chapter 1.

⚠ Pick a booking you finished. Try to write down, from memory, what the property looked like on the fourth day. What was on the kitchen worktop. Whether the ceiling was clean. Whether anything smelled. Whether the back door was locked at nine in the evening.

⚠ You will not be able to, and neither would anybody.

⚠ That blank is the thing this book is about, and it is entirely fixable with about four minutes a day.

---

©2026 James Henderson / https://localhandyman.work

Chapter 1 — Understanding the Industry and What People Actually Buy

Nobody buys presence - they buy occupancy, animal care, and the assurance that somebody was paying attention.

---

1.1 Who actually buys

⚠ The assumed customer is somebody going on holiday who wants their plants watered. That customer exists, is the least valuable, and is the one most likely to use a free exchange platform instead.

⚠ The people who actually sustain a house sitting business fall into six groups, and their reasons for paying are entirely different.

Pet owners who will not use a kennel or a cattery. By a distance the largest paying segment. The animal is the reason, the house is where the animal lives, and the price is set by what the alternative costs. These clients are loyal, they book far ahead, and they will pay properly.

⚠ People with a property that cannot be left. An old house, a large garden, a swimming pool, a rural place with a septic tank or a well, a listed building, a property with a history of problems. Their fear is not burglary. It is a burst pipe in February with nobody there for three weeks.

People whose insurance or mortgage terms make an empty property a problem. Unoccupied-property conditions exist and many owners discover them the first time they plan a long absence. Whether they apply to a given policy is a question for the insurer, and the demand they create is real.

⚠ Frequent travellers and people who work away. Recurring, predictable, and the closest thing to a subscription this trade offers.

People in a life event. A hospital admission, a spell in rehabilitation, a family emergency abroad, an extended stay with a relative. Unplanned, urgent, often emotional, and frequently arranged by somebody other than the owner.

Second-home and holiday-let owners. Between guests, over a closed season, or during works. Commercial in tone, scheduled, and under-served.

⚠ Note that only the first group is about pets and only one group is about holidays. The stereotype covers a small part of the market.

---

1.2 What they are really buying

⚠ Nobody buys house sitting. Four things are being bought, and knowing which one is in front of you determines your price and your service.

They are buying an animal being properly looked after in its own home. For the largest paying group this is the entire product. The house is incidental; the cat that will not eat in a cattery is the point.

⚠ They are buying somebody being there when something goes wrong. Not preventing it — being there. A pipe that bursts with somebody in the house is a mopped floor and a plumber. The same pipe in an empty house for three weeks is a ceiling, a floor, and a claim.

⚠ They are buying the ability to stop thinking about it. This is the one owners find hardest to articulate and the one they value most. Somebody going away for a month does not want a daily worry about whether the freezer has failed. What they are paying for is the removal of a background anxiety, which is why the daily update matters far more than it looks. Chapter 24.

They are buying the appearance of occupation. Lights that change, bins that go out, post that does not pile up behind the door, a car on the drive. Whether this deters anybody is not something this book will claim; that it is bought is a fact.

⚠ And underneath all four: they are buying somebody they can trust alone in their home. That is a bigger ask than any other domestic service makes, and how you handle it is the whole of Chapter 9.

---

1.3 Why the market is bad at this

⚠ The market is served by four kinds of provider and each fails in a way that creates your opening.

Exchange and platform sitting. Sitters stay free in exchange for looking after the property, usually mediated by a subscription platform. Enormous, popular, and structurally limited: the sitter is choosing the property, availability is unpredictable, cancellations are common, and the relationship carries no fee and therefore limited obligation.

⚠ Its real weakness is continuity. The owner gets a different person each time, which means the property file is rebuilt from scratch every booking and nobody accumulates knowledge of the house.

Agencies. Vetted, insured, structured, expensive, and often unavailable at short notice or in less populated areas. They also rotate sitters.

Friends, family, and neighbours. Free, trusted, and unstructured. No agreement, no record, no insurance position, and — critically — no ability to complain without damaging a relationship. This is the largest provider in the market and the one nobody counts.

⚠ Kennels and catteries for the animal, plus a neighbour for the post. The default, and the thing your pet-owning clients are trying to avoid.

⚠ Every one of those fails in the same place: nobody holds a maintained record of a specific property, and nobody produces an account of the period they were responsible for. The exchange sitter has no reason to. The agency's sitter is different next time. The neighbour would not think of it.

That gap is the whole competitive position. A named person who returns to the same property, holds a property file that gets better each booking, and produces a dated daily record of every day they held the keys.

⚠ Chapter 17 builds the file. Chapter 24 builds the record. Together they are the reason a client books you rather than the alternative.

---

1.4 Where demand concentrates

⚠ Demand in this trade is unusually lumpy, and the shape of it determines whether the business is viable.

By season. School holidays, Christmas and New Year, and the summer weeks are the peaks. Half-terms and long weekends fill the gaps. February and November are thin.

⚠ The consequence is severe: peak weeks are booked six to twelve months ahead, everybody wants the same fortnight, and your annual income is largely determined by how well you manage a calendar rather than by how good you are. Chapter 30.

By duration. A week to a fortnight is the standard shape. Long bookings — a month or more — are disproportionately valuable because the fixed cost of intake, handover, and departure is amortised. Two-night bookings are frequently uneconomic once travel and setup are counted. Chapter 28.

By property type. Rural, large, or complicated properties pay more and demand more. Flats and small modern houses are simpler and cheaper.

⚠ By pets. The single biggest driver of both price and complexity. A booking with three animals, one of which needs medication, is a different job from an empty flat, and it should not share a rate.

By geography. Less constrained than most trades, because a booking is a stay rather than a route — you can travel to a booking in a way you cannot travel to a weekly shop. That widens your market considerably and it introduces travel cost and time that most sitters never price. Chapter 14.

⚠ And by relationship. The most valuable demand is repeat: the same family, twice a year, for years. Chapter 27 argues that a business built on repeat bookings and one built on new enquiries are different businesses with different economics.

---

1.5 What kills these businesses

⚠ Five things, and only one of them is being bad at the job.

⚠ An unrecorded period. The Sacha failure. Something is discovered, the question is when it started, and there is no answer for the days you were there. It does not require you to have done anything wrong and it can still end the relationship and trigger a claim.

⚠ An access failure. A lost key, a shared code, a door left unlocked, an alarm not set, somebody else in the house. The highest-cost failure in the trade, the one owners find least forgivable, and the one sitters are most casual about. Chapter 6 and Chapter 18.

A boundary failure. A guest, a partner staying, using the car, a room you had no reason to be in. Almost always innocent, almost always undisclosed, and almost always discovered — because houses have neighbours, cameras, and a smell of somebody else's cooking. Chapter 5.

⚠ An animal incident. A pet that escapes, gets ill, or dies. Sometimes preventable and frequently not, and it is the failure with the most emotional weight in the entire trade. Chapter 21 is twelve sections for that reason.

And the one that is not a disaster but a slow strangle: an empty calendar. This is a business where income is bookings multiplied by rate, and where the peaks are contested and the troughs are dead. A sitter with excellent reviews and forty per cent occupancy is not making a living.

⚠ Note that four of those five have nothing to do with whether you look after houses well, and three of them are solved by writing something down.

---

1.6 How AI Helps

Mapping which buyer groups exist near you. Feed it your area's characteristics — housing stock, rural or urban, demographics, distance from an airport — and ask which of the six groups are likely under-served. A hypothesis list to go and test.

Building the property file questions. The questions that produce a usable file are far more numerous than anybody expects, and this is the single best use of AI in the book. Chapter 17.

Turning a handover conversation into a written file. Excellent.

Modelling a booking calendar. Occupancy, peaks, travel time between bookings, and what a realistic year looks like.

⚠ What it cannot do: tell you what local demand is, what anybody charges, whether a property is safe, whether an animal is well, or what any insurance covers. Every one of those is a named human source, dated in Resource 3.

---

Prompt 1 — Buyer Group Mapping

"I am considering starting a house sitting business. My area and situation: [describe location type, housing stock, rural or urban, proximity to airports or transport, local demographics, and whether you can travel to bookings]. Using these six buyer groups — pet owners avoiding kennels, owners of properties that cannot be left empty, owners with unoccupied-property insurance concerns, frequent travellers, people in an unplanned life event, and second-home or holiday-let owners — estimate for each: likely demand in my area, what they would specifically want, what they would compare me to, and what evidence I could gather in one week to test whether that demand is real. Mark clearly which estimates are guesses. Do not invent statistics or state anything about insurance rules."

Prompt 2 — What They Are Actually Buying

"Here are three descriptions of prospective house sitting clients, in their own words: [paste or describe what each said]. For each, identify which of four things they are actually buying — an animal looked after in its own home, somebody present when something goes wrong, the removal of a background worry, or the appearance of occupation — and what evidence in their words points to it. Then tell me, for each, what would make them consider the booking a failure even if nothing went wrong, and what single question I should ask at intake to avoid that. Flag anything suggesting they expect security, repairs, or veterinary care from me."

Prompt 3 — Demand and Calendar Shape

"Help me understand how house sitting demand concentrates so I can plan a year. Cover: seasonal peaks and troughs; typical booking durations and why very short bookings are often uneconomic; how property type and pets drive both price and complexity; how far ahead peak weeks are typically booked; and how travel to a booking widens the market but adds cost. Present it as a planning framework with every assumption stated explicitly, not as data. Then tell me what occupancy rate I would need across a year to make this viable at a given day rate, showing the arithmetic and leaving the rate blank."

Prompt 4 — Failure Mode Audit

"Here is how I currently sit houses, or how I plan to: [describe your arrangement — how you learn about a property, what you record during a booking, how keys and codes are handled, what you do about pets, and what happens in an emergency]. Audit it against five failure modes: an unrecorded period where nobody can say when something started; an access failure involving keys, codes, alarms, or somebody else entering; a boundary failure involving guests or use of the owner's things; an animal incident; and an under-filled calendar. For each, tell me exactly where I am exposed, the realistic worst case, and the one change that would most reduce it. Be blunt."

---

1.7 Common Mistakes

⚠ Assuming the customer is somebody on holiday with plants. The smallest and least valuable segment.

⚠ Treating a pet booking as a house booking with an animal attached. It is a different job with a different risk profile and it should not share a rate.

Competing with free exchange platforms on price. You will lose. Compete on continuity, the property file, and the daily record.

⚠ Ignoring the calendar as the primary constraint. Excellent reviews and forty per cent occupancy is not a living.

Accepting very short bookings at a day rate. Once travel, handover, and departure are counted they frequently lose money.

⚠ Believing that nothing going wrong is the measure. The measure is whether you can say what happened.

---

1.8 Safety, Legal, Ethical, and Professional Considerations

⚠ Before you accept a single paid booking, get answers from named human beings to a specific list.

⚠ What your own insurance covers while you are staying in and responsible for somebody else's property — and, separately, what the owner's insurance covers while they are away and somebody else is in the house. These are two different questions and both need asking.

⚠ Whether staying in a property for an extended period creates any occupation, tenancy, or licence position where you are. This is a genuine legal question with a jurisdiction-specific answer and it is the one nobody asks.

⚠ What obligations you take on by accepting responsibility for an animal, and what your position is if one is injured, escapes, or dies.

⚠ What obligations attach to holding keys, alarm codes, and access to a property — including what happens if you lose them.

What data protection obligations apply to the property files, absence dates, and access details you hold. Chapter 16.

⚠ And whether any background check is available to you, expected of you, or required for this work in your area.

Every answer goes into Resource 3 with a name and a date.

Ethically, this trade gives you unsupervised access to how people actually live — their post, their finances on the hall table, their medicine cabinet, their arguments on the answering machine. The professional standard is simple: you see it, you do not look at it, you never repeat it, and you record only what the job requires.

---

1.9 Do This Now

1. Start Resource 3 — Dated Source and Rule Register with the seven questions from Section 1.8, unanswered, dated today.
2. Ring your insurer and ask the two insurance questions — yours and the owner's — before your next booking.
3. Complete Resource 1 — Claim and Credibility Register with every claim you currently make about yourself.
4. Begin Resource 7 — Competitor Research Worksheet covering platforms, agencies, kennels, and the invisible neighbours-and-family segment.
5. Sketch a year's calendar and mark the peaks. Note how far ahead each would need to be booked.
6. Decide, provisionally, whether you take pet bookings. It is the single biggest choice in this business. Resource 6.

---

1.10 Chapter Summary

The stereotype — a holidaymaker with plants — is the smallest and least valuable part of this market. The paying customers are pet owners avoiding kennels, owners of properties that cannot be left, owners with unoccupied-property concerns, frequent travellers, people in an unplanned life event, and second-home owners.

They are buying an animal looked after in its own home, somebody present when something goes wrong, the removal of a background worry, and the appearance of occupation — and underneath all four, somebody they can trust alone in their home, which is a bigger ask than any other domestic service makes.

Every alternative provider fails in the same place: nobody holds a maintained record of a specific property, and nobody produces an account of the period they were responsible for. Exchange sitters have no reason to; agencies rotate; neighbours would not think of it.

Demand is lumpy by season, duration, property, pets, and relationship, and the calendar is the primary constraint on income rather than skill.

Five things kill these businesses: an unrecorded period, an access failure, a boundary failure, an animal incident, and an empty calendar. Four of the five have nothing to do with looking after houses well, and three are solved by writing something down.

---

©2026 James Henderson / https://localhandyman.work