Disclosure: As an Amazon Associate, umtsmarttool.com earns from qualifying purchases. Links to Amazon on this page are affiliate links; they cost you nothing extra. Prices and availability change often, so check the current Amazon listing before you buy.
Where to buy
The three alarms the steps below most often land on — each verified to its manufacturer and model: the dual-sensor anchor, the First Alert SA3210; the slim photoelectric for living spaces, the First Alert PR710; and the combination unit with a carbon monoxide display, the X-Sense SC07. Read the full guide ↓
How to Choose a Smoke Detector: a Seven-Step Walkthrough

Key Takeaways
- Count the positions from your floor plan before you look at a single product
- Decide carbon monoxide cover by one question: does anything burn fuel in or beside the home?
- Pick sensing by room: dual or photoelectric in bedrooms, calm sensing near the kitchen
- Choose sealed ten-year power unless your household genuinely keeps a battery rhythm
- Interconnection is a brand-family decision — make it once, on paper
- Check the manufacture date on delivery; the decade runs from that date, not from purchase
- Write the log entry the day you fit each unit: model, room, date
- Choose per room, not per house — mixed specifications across positions are correct, not indecisive
- When two steps conflict, the earlier step wins: positions before products, hazards before features
- The manufacture date on the delivered unit outranks every claim on the listing
Choosing a smoke detector is a procedure, not a taste: seven decisions in a fixed order, each one narrowing the next, until the product names itself. This walkthrough takes them in sequence, with the question each step answers, the options stated plainly, and the safety notes flagged where they belong. Work it with a floor plan in hand — a sketch is enough — and a notebook for the log you will start in step seven. Readers who want the theory behind the steps will find it in Smoke Detector Buying Guide Smoke Detector Buying Guide; this page is the practice. One convention before starting: whenever a step offers a judgement call, this walkthrough names the default for an undecided household, because a procedure that cannot decide for the undecided is only half a procedure. Defaults are marked as defaults; a household that knows its own answer should take it — the procedure’s loyalty is to the house, not to its own suggestions. Where a default is overridden, the log should say so in a word or two; future choosers deserve to know which decisions were deliberate.
In this guide
Step One: How Many Alarms Does Your Floor Plan Demand?
Walk the house on paper, doors in their night positions. Mark one position inside every bedroom, one outside the sleeping area, and one on every level including the basement. That count is the shopping list’s length, and it is decided before any product enters the conversation. Safety note: resist the economy of the single hallway unit — a closed bedroom door is precisely the barrier the in-room position exists to defeat, and no premium model compensates for a missing position. Include the awkward rooms honestly: the nursery-to-be, the lodger’s room, the converted loft, the basement with the freezer and the boiler. Rooms change use faster than ceilings change alarms, and a position counted generously today is a renovation that needs no second thought tomorrow. If the count surprises you — most households guess low — trust the floor plan over the guess; the plan is the only witness that has actually been in every room.
Done well, step one produces a number and a sketch, nothing more — and produces them before any shop, site or shortlist is opened. Households that find the step tedious should note its leverage: every later step multiplies by this number, so an error here is the only error the procedure cannot correct downstream. Keep the sketch even after fitting; it is the map the annual test walks, room by room, for the whole decade.
Step Two: Does Your Home Need Carbon Monoxide Cover Too?
Answer one question: does anything burn fuel in or beside the home — boiler, gas hob, wood stove, fireplace, attached garage? If yes, combination smoke-and-CO units (listed to UL 2034 as well as UL 217) take at least the hallway positions, and the fuel-adjacent level deserves one too. If the home is genuinely all-electric with no garage, smoke-only units are a defensible saving. Safety note: CO is undetectable by human senses; when in doubt, cover the hallway — doubt is information. Extend the question through the decade the alarms will serve: a planned stove, a generator bought for outages, a garage conversion, a lodger with a gas heater — the ceiling is being chosen for the house’s future as much as its present, and combination cover bought early is the cheapest version of the decision.
Done well, step two is a single sentence in the log — “fuel: boiler + hob + garage; combination at landing, hall, basement” — that step six will simply obey. If the sentence cannot be written, the house’s appliances need listing first; guessing at this step is how fuel homes end up with smoke-only ceilings. Revisit the sentence whenever an appliance changes; it is the shortest document in the procedure and the one with the longest consequences.
Step Three: Which Sensing Belongs in Which Room?
Bedrooms fear the smouldering fire: choose dual-sensor (photoelectric plus ionisation) or photoelectric units there. Near the kitchen and bathrooms, calm matters: photoelectric or multi-criteria designs, placed at least ten feet from cooking appliances, resist the steam-and-aerosol false alarms that train households to disable devices. Living spaces follow the kitchen rule if cooking is open-plan, the bedroom rule otherwise. Safety note: never solve a nuisance problem by removing a battery or a unit — reposition, respecify, or use the hush; a gap is not a solution. Write the sensing assignment on the floor-plan sketch itself, room by room, so the shopping list carries it: “dual”, “photo”, “combo + display”. The sketch becomes the order, the delivery check and the log’s first draft in a single sheet of paper — the most efficient document in the whole procedure.
Done well, step three leaves the sketch annotated and the household in agreement about the kitchen position — historically the argument this step exists to settle early, on paper, rather than at midnight after the third false alarm. The annotations also future-proof the replacements: in ten years, nobody will remember why the hallway unit differs from the bedroom units, but the sketch will say.
Step Four: Sealed, Replaceable or Hardwired Power?
Ask the household one honest question: can anyone name the month the current batteries were last changed? If not, choose sealed ten-year units — power and sensor retire together, and the borrowed-battery failure mode disappears. Replaceable-cell units suit disciplined households and some specialist positions. Hardwired homes should replace like-for-like within the existing system’s brand family, keeping the wired interconnect intact; changing a wired system wholesale is an electrician’s conversation, not a weekend improvisation. Treat this step as a judgement about people rather than products: the best power system is the one whose maintenance your household will actually perform in year six, tired, in February — and for most households that system is the one that asks for nothing.
Done well, step four is decided by the household’s memory rather than its intentions — the honest answer about batteries, given once, protects the decade better than a virtuous answer repeated annually and performed never. Households that choose replaceable cells anyway should write the battery date into the log beside the fitting date, giving the habit an artefact to attach to.
Step Five: Should the Alarms Sound Together?
If the house has more than one floor, closed bedroom doors, or a basement, interconnection earns its place: one event sounds everywhere. Choose the family deliberately — wired where the wiring exists, wire-free radio networks or a connected platform for retrofits — knowing that brands do not mix and the first unit commits the decade. Single-level, open-plan homes may honestly skip this step and spend the money completing step one’s count instead. Record the family decision at the top of the log in capital letters if necessary: every future purchase, gift and replacement reads it first, and the ten minutes spent choosing the family well is repaid at every subsequent step six.
Done well, step five ends with a family name in the log and a short list of positions marked “networked” — the two facts every later purchase, gift and replacement will consult before anything else. If the household splits the ceiling between a network and standalone anchors, the sketch should show the boundary explicitly, so a future buyer of one unit knows which world it must join.
Step Six: How Do You Read the Listing Before You Commit?
Run the ninety-second drill on every candidate. Standards line first: UL 217 named, UL 2034 for combination units — vague or absent, move on. Sensing in nouns, not adjectives. Power stated in years. Hush and end-of-life signal present. Then the model check: the name on the listing must match the manufacturer’s own page exactly, because near-namesake models with different sensing are this category’s quietest trap. Safety note: buy through channels with easy returns — step seven explains why the returns policy is part of the safety equipment. Run the drill out loud if a partner is choosing with you; the standards line and the sensing nouns, spoken, expose vagueness that silent reading forgives, and agreement reached at the listing is an argument saved at the ceiling.
Done well, step six produces a basket whose every line can be defended in one sentence — the standard, the sensing, the years — and deletes more listings than it keeps, which is the drill working as designed. Save the drill’s rejects with a one-word reason in the log; at replacement time, the rejected list is a map of traps already mapped.
Step Seven: What Happens on Delivery Day?
Photograph the manufacture date before fitting anything; if the unit arrives deep into its decade, return it — the decade is the product. Fit per the placement rules (ceiling centre or at least four inches from a wall; ten feet from the cooker), run the test with the household listening from the furthest bedroom, and write the log entry immediately: model, room, manufacture date, replacement year. The choosing is finished only when the log exists; an unlogged alarm is a future mystery. Add one line to the log that step lists rarely mention: where the receipt lives. Warranty questions, insurance questions and the next decade’s shopping all begin at that line, and it costs eleven seconds to write while the box is still open.
Done well, step seven is the only step that touches a ladder — and it ends the choosing so completely that the household’s next decision about alarms is a decade away, pre-written in the log. Photograph the finished ceiling too; the picture, mundane as it seems, is the fastest orientation document the next fitting day will ever have.
Troubleshooting the Choosing Itself
Procedures fail in predictable places; these are this one’s, with their resolutions stated in the same order the steps would catch them. Read them before starting if the house is unusual — a split-level plan, a granny annexe, a workshop with its own stove — because unusual houses fail at unusual steps, and forewarned is genuinely forearmed here:
Problem: two shortlists disagree about the “best” alarm. Resolution: they are weighting different rooms — return to step three and choose per position rather than per house; mixed specifications across rooms are normal and correct. If the disagreement survives the sketch, it is usually about step five’s family, and should be settled there, once, rather than re-fought at every position.
Problem: the count exceeds the budget. Resolution: phase by risk — bedrooms and the sleeping-area hallway first, remaining levels on a dated plan in the log; never solve it by deleting a bedroom position to afford a premium model elsewhere. A dated phase-two in the log is a commitment the household can be held to — by itself, at the next payday — in a way a mental note cannot.
Problem: the home has an existing wired system of unknown brand. Resolution: photograph a unit’s label, identify the family from the manufacturer’s documentation, and replace like-for-like; if the family is discontinued, treat it as a whole-system decision with professional advice. Do not let one failed wired unit talk you into a mixed ceiling of strangers — the interconnect is the system’s value, and it survives only intact.
Problem: a household member wants the alarm moved after a false alarm. Resolution: fix the cause in order — distance from the source, sensing type at that position, hush discipline — and move the unit only within the placement rules; the bedroom position itself is not negotiable.
Problem: every listing claims ten years, so the claim feels meaningless. Resolution: it is — as marketing. The operative fact is the manufacture date on the delivered unit (step seven); judge sellers by their returns policy on stale stock, and the claim by the box, never the banner. Price per remaining year, computed from the photographed date, is the only honest comparison between a fresh unit and a discounted stale one.
Problem: the household cannot agree between connected and conventional. Resolution: split the ceiling deliberately — one connected unit at the hallway for reporting and self-testing, conventional sealed anchors in the bedrooms — and record the hybrid as the family decision in step five’s log line. Hybrids chosen on paper behave like systems; hybrids that happen by drift behave like arguments, so write the intention down.
Gear That Makes This Easier
Three verified units cover most outcomes of the steps above. The First Alert SA3210 is the step-three answer for bedrooms: dual sensing, sealed decade, end-of-life signal. The First Alert PR710 is the calm, slim answer for living spaces and kitchen-adjacent positions. The X-Sense SC07 is step two made hardware: combination cover with a CO display for fuel-burning homes. The ranked field with full reviews, once the steps have produced your shortlist, is at Best Smoke Detector For Home Use Under $50 Best Smoke Detector For Home Use Under $50. Treat the gear as worked examples rather than commandments: any unit that passes step six’s drill and matches your sketch’s per-room assignments serves the procedure equally — the steps are the product this page is really recommending, and they are free. Notice what the three examples share beyond verification: each states its standard, its sensing and its years without being asked twice — the surface signature of a listing written by a manufacturer that expects to be checked. Keep the drill’s printout or screenshots with the log if you like; at replacement time they show the next chooser exactly what “passing” looked like.
Mistakes to Avoid
Choosing the model before the count. Step order is the whole method; a brilliant product at a missing position protects nothing. One specification for every room. Bedrooms, kitchens and hallways have different physics; the steps assign accordingly, and the sketch records the differences. Reading “10-year” from the purchase date. It runs from manufacture — step seven’s photograph is the correction, and a stale bargain is a shorter product at a friendlier price per box and a worse one per year. Mixing interconnect brands. Families do not interoperate; step five is a single, recorded decision, made before the first unit is bought rather than discovered at the third. Skipping the log. Every future decision — testing, replacing, selling the house — reads from it; the five minutes it costs are the cheapest step of the seven. Treating the choosing as finished at checkout. Delivery day is a step, not an epilogue: dates, fitting and the first full-house test are where the procedure’s promises are either kept or quietly lost.
A Worked Example: the Whole Procedure in One House
Watch the steps run on a typical property — a three-bedroom semi with a gas boiler, an open-plan kitchen-diner and a basement utility room. Step one counts six positions: three bedrooms, the landing, the ground-floor hallway, the basement. Step two answers yes to fuel (boiler, hob, utility room below the bedrooms), so combination units take the landing, the hallway and the basement positions. Step three assigns dual-sensor units to the bedrooms, a photoelectric unit’s calm to the kitchen-adjacent hallway, and combination displays where the fuel story lives. Step four: nobody in the house can date a battery change, so everything is sealed. Step five: two floors and closed doors argue for interconnection; the household chooses a single brand family for the networked positions and records the choice. Step six runs the drill on the shortlist — standards, nouns, years, hush, end-of-life — and deletes one attractive listing whose sensing was adjectives only. Step seven: delivery day, dates photographed (all fresh), six log entries written, one test heard from the furthest bedroom with its door shut. Total choosing time, just under an hour; total decisions, seven; the ceiling then asks for fifteen minutes a year until the log says otherwise. Any house can be run through the same narration — write yours before you shop, and the shopping becomes transcription. Notice, in the narration, how late the products appear: five steps pass before a brand is named, which is the procedure’s whole discipline — the house speaks first, the catalogue answers. Try narrating a flat instead — two bedrooms, electric everything, one floor — and watch the procedure shorten itself: four positions, no combination cover, standalone units honestly sufficient, the whole choosing done in twenty minutes. The steps scale because they ask questions rather than reciting answers. Scale works upward too: a landlord with six flats runs the same narration six times, and the only step that changes character is the log, which becomes a portfolio record rather than a household one.
Choosing for a House You Do Not Live In
The procedure adapts, with three emphases, when the ceiling belongs to a parent, a tenant or a child leaving home. Step four hardens: sealed power, no exceptions, because you cannot audit another household’s battery habits from a distance. Step three leans calm: nuisance behaviour in someone else’s home cannot be supervised by you, so photoelectric peace near kitchens is worth more than marginal sensing breadth. Step seven doubles: the log lives in both houses — the fitted home’s copy with the property papers, yours with the visit notes — and the test is demonstrated together, once, so the sound is familiar rather than frightening. Landlords run the same adaptation with the duties reversed: the log is a compliance record, the fitting is documented at tenancy start, and step two’s fuel question is answered from the property’s appliances, not the tenant’s habits. The steps do not change; their weights do, and weighting is what choosing for others honestly means. One more adaptation belongs to distance itself: record the exact model where the recipient will find it, because the replacement in year ten will be bought by someone who was not in the room for step six — the log is how the choosing survives its chooser. Where the recipient is a landlord’s tenant rather than family, the same steps apply with the duties documented rather than assumed — the choosing-for-others section above weights them. Fit for others on a visit you would have made anyway; the procedure should ride on care that already exists, not demand new occasions.
Choosing Under a Deadline: the New-Home Weekend
Moves compress the procedure, so here is the deadline version, honestly labelled. Day one, before the furniture: walk the empty house with the sketch — empty rooms make step one faster and more accurate than it ever is again — and audit whatever the previous owners left, dates first. Order that evening against the sketch, choosing sealed units throughout (a new household has no battery rhythm yet, by definition) and combination cover wherever step two’s fuel answer is yes. Fitting day: positions per the sketch, dates photographed, the log started in the moving paperwork folder where it will actually be found. The deadline version skips nothing; it sequences everything into the one weekend when the ceilings are empty, the ladder is already out and the household is thinking about the house as a system. Buyers who inherit a fitted system mid-decade should run the same weekend in reverse: audit first, replace only what the audit condemns, and let the log record the inheritance. Photograph every inherited date before touching anything — the previous owners’ ceiling is evidence about the house’s maintenance culture, and the photographs preserve it whichever way the evidence points. A move is also the one moment when replacing everything at once is cheap in attention: the household is already living out of boxes, and a single fitting day beats five separate ones.
What Good Choosing Feels Like a Year Later
The procedure’s success is measurable, a year on, by absences: no false-alarm stories, no battery drawer, no orphaned unit that will not join anything, no mystery chirp unresolvable because nobody knows a date. The household tests on rhythm, vacuums in spring, and thinks about its ceiling approximately never — which is the state the seven steps were designed to produce. When a step was skipped, the year reports that too, specifically: the skipped count announces itself as the room that feels unprotected, the skipped family decision as the unit that will not pair, the skipped log as the argument about when anything was fitted. Read the year’s report honestly at the first annual test, correct the one skipped step it names, and the procedure has then done its full work — not merely choosing the alarms, but teaching the household how its own ceiling thinks. A household that has run the steps once owns a skill, not just a system: every future house, relative’s flat and rental audit runs faster, because the questions are already in the right order in somebody’s head. The year’s report is also the moment to update the sketch for life changes — the nursery that became a bedroom, the study that became a bedsit — because the count is a living number, and the procedure’s last kindness is reminding the household of that annually, in writing.
If You Only Have Ten Minutes
The compressed procedure, for the reader whose ceiling cannot wait for the full hour — with the honest warning that compression spends accuracy. Minute one: count bedrooms, add the hallway and each level (step one, approximate). Minutes two to three: answer the fuel question and mark the hallway for a combination unit if yes (step two). Minutes four to five: assign dual or photoelectric to bedrooms, photoelectric calm near the kitchen (step three). Minute six: sealed, everywhere (step four’s default). Minute seven: if floors and doors demand it, pick the family whose name you already know from the sketch’s likely units (step five, provisional). Minutes eight to nine: run the drill on exactly one listing per position type (step six). Minute ten: order, and put the full procedure — sketch, dates, log — in the diary for delivery day (step seven, uncompressed). The ten-minute version buys the right shape of ceiling; the hour version buys the right ceiling. Where the ten minutes were spent, schedule the hour within the month and correct what it finds — the procedure forgives compression, provided compression is confessed in the log. The likeliest corrections, for the record: an undercount in a split-level plan, a fuel appliance forgotten in the garage, and a kitchen position that wants moving ten feet rather than enduring — all cheap in month one, all expensive in year three.
The Seven Steps on One Card
- 1 Count — bedrooms, hallways, levels, from the plan
- 2 Hazards — fuel anywhere → combination at the hallways
- 3 Sensing — dual/photo for bedrooms, calm near kitchens
- 4 Power — sealed decade unless the household proves otherwise
- 5 Network — one family, chosen once, written down
- 6 Listing — standards, nouns, years, hush, end-of-life
- 7 Log — date photographed, entry written, receipt filed
Choosing With a Partner: Dividing the Steps
Two-person households choose better when the procedure is divided by temperament rather than duplicated by argument. Give steps one and two — the sketch and the fuel question — to the partner who knows the house’s fabric; give step six, the listing drill, to the partner who reads specifications coldly; steps four and five, the habit and family judgements, belong to both, over the sketch, because they encode how the household actually lives. Step seven is a fitting party, not a chore: one on the ladder, one on the log, the test heard together from the furthest room. The division matters because the classic choosing failure is not ignorance but a single enthusiast optimising one step — usually the product step — while the count and the log go unowned. A procedure with named owners for its unglamorous steps is a procedure that survives contact with a real weekend. Solo choosers should borrow the discipline differently: run the steps on two separate sittings, sketch first and listings second, so the catalogue never gets to argue with the floor plan while both are open. A night’s sleep between the sittings is the solo chooser’s substitute for a second opinion.
The Steps at Work on Three Candidate Listings
Abstract drills convince nobody, so run step six over the three shapes of listing a chooser actually meets. The complete listing: standards named with editions, sensing in nouns, power in years, hush and end-of-life stated, model identical on the manufacturer’s page — it passes in ninety seconds and earns the basket; most of this cluster’s linked picks read this way because they were chosen by the drill. The enthusiastic listing: adjectives in place of principles, a decade claimed without a power story, features listed generously and standards whispered — it fails at the nouns, and the correct response is not suspicion of the brand but a trip to the manufacturer’s own page, where the same model either states its facts or ends the conversation. The bargain listing: a sound model at a startling price — the drill passes it, and step seven carries the verdict: the photographed date decides whether the bargain is a fresh unit discounted or an old unit expensively cheap. Three listings, three different steps delivering the decision — which is why the procedure has seven steps rather than one clever question: each listing fails, or passes, at its own weakest point, and the steps exist to find that point before the ceiling does. Practise the drill on a listing you have already decided against; calibration on a known reject makes the live runs faster and colder. Then file the three shapes in memory — complete, enthusiastic, bargain — because recognising the shape is most of the drill.
A Note on Second-Hand, Inherited and Gifted Units
Units arrive by routes other than the procedure, and the steps adapt rather than collapse. An inherited unit (fitted by previous owners, or found in a drawer) enters at step seven in reverse: photograph its date first — if the decade is spent or nearly spent, the unit is recycling, whatever its condition appears to be, because sensing ages invisibly. A gifted unit enters at step five: does it join the household’s family, or stand alone honestly? Either answer is acceptable; the unacceptable outcome is the unexamined assumption that it networks. A second-hand unit enters at step six with the drill doubled: standards and model verified against the manufacturer’s page, date demanded before money moves, and no purchase at any discount when the date is hidden — the discount on an alarm is paid in years, and years are the product. The procedure’s generosity is that it has a door for every route in; its firmness is that every door leads to the same log. Units that fail their door’s checks should be recycled promptly rather than stored “just in case” — a drawer of doubtful alarms is a liability dressed as thrift. The same firmness applies to the half-expired unit “kept for the garage”: garages adjoin houses, and doubtful protection there is doubtful protection, full stop.
The Procedure in a Rental, Briefly
Renters run the steps with two substitutions. Step one counts the landlord’s provision as well as the requirement, and the sketch marks which positions are whose — because the gaps a renter fills are usually the landlord’s undone duties, and the sketch is the evidence. Step four defaults harder to sealed: a tenancy may end mid-decade, and the units a renter fits should need nothing from the next occupant to stay honest. Everything else — hazards, sensing, the drill, the log — runs unchanged, with the log kept in the renter’s own papers rather than the property’s. The deeper treatment, including what to put in writing to a landlord whose provision fails the audit, lives in the cluster’s rental guides; the procedure here is deliberately the same one owners use, because the physics of smoke does not check tenure. Renters who move often should keep the sketch template itself; the next flat’s procedure then starts from a practised form rather than a blank page. And a renter’s own sealed units, bought for a gap the landlord left, travel to the next home — one more reason the log, not the property, should hold their record.
Why the Order Matters: the Steps as a Dependency Chain
Each step consumes the previous step’s output, which is why reordering breaks the method rather than merely rearranging it. The count feeds the budget; the hazards feed the specification; the specification feeds the drill; the family decision constrains the basket; the log preserves all of it. Run the drill first — the internet’s default order — and the chooser falls for a product before knowing how many positions exist, which hazard mix they carry, or whether the beloved unit can join the network the house needs. The product then dictates the ceiling backwards, and the gaps appear exactly where the skipped steps would have spoken: the uncounted basement, the unconsidered boiler, the unjoinable third unit. Procedures feel pedantic precisely at the point where they are preventing the expensive version of the same afternoon. If the order chafes, compress the steps — the ten-minute version does — but never invert them; a fast chain holds, an inverted one is just shopping with extra confidence. The chain also explains the page’s own shape: products appear late here for the same reason they should appear late in your choosing. Readers in a hurry can still use the chain backwards as a diagnostic — start from the symptom (a chirp, a gap, an orphan) and walk up the steps to find which one was skipped.
Frequently Asked Questions
How long does choosing properly take?
About an hour with the floor-plan sketch, most of it in steps one and two. The steps are deliberately front-loaded: an hour of choosing buys a decade without re-deciding. Households that rush the hour reliably spend it later anyway — at the ceiling, with the wrong unit in hand, re-running step three in worse light. Budget the hour honestly: forty minutes at the table with the sketch, twenty at the screen with the drill, and fitting day kept separate so neither is hurried by the other.
Can I choose different brands for different rooms?
Yes, where units are standalone. The moment interconnection enters (step five), the networked positions must share a family — plan which positions network before mixing brands elsewhere. Note the decision on the sketch; a future gift or replacement bought in ignorance of it creates the orphaned unit the troubleshooting section warns about.
What if my landlord already fitted alarms?
Run the same steps as an audit: count, hazards, dates, tests. Fill genuine gaps with your own sealed battery units, and put failures to the landlord in writing — the duty is usually theirs, the sleep is yours. Keep your audit notes with the tenancy agreement; if the provision is ever disputed, the dated audit is the document that settles it.
Is photoelectric or dual better for a child’s bedroom?
Both answer the smoulder scenario; dual adds flaming-fire speed. Choose dual where the budget allows, photoelectric where calm near an en-suite matters more — and either over ionisation-only. For a child who is a light sleeper, add the hush to the specification and rehearse the alarm sound together once, in daylight, so the night version is familiar rather than terrifying. Note the rehearsal in the log; children outgrow the fear, but the note reminds the next fitting day to rehearse with the next child.
What is the most commonly skipped step?
Step one — households shop first and count later, then discover the count at the ceiling with the wrong quantity in hand. Step seven runs it close: the log feels optional on a busy fitting day, right up until the first chirp asks a question only the log can answer. Both skips share a cause — they are the steps whose benefits arrive years later, which is precisely why a written procedure, rather than enthusiasm, has to carry them.
Do the steps work in a shop, or only online?
Both — the sketch and the card travel on paper or a phone, and step six’s drill reads a box as happily as a listing. In a shop, add one check: the physical unit’s date is visible before purchase, which is the one advantage the shelf holds over the screen. Staff can be a step-six resource too — ask which models get returned and why, and treat the answer as field data rather than sales talk.
When should I re-run these steps?
At every replacement decade, at any renovation, and whenever a fuel appliance arrives or leaves. The steps are cheap to re-run and the house they describe keeps changing. Keep the old sketch with the log; comparing the two is the fastest history of the house’s risk that anyone will ever write.
The Procedure, Memorised in One Breath
For the reader who will run this from memory in a shop aisle: Count the rooms. Ask about fuel. Match the sensing to the room. Seal the power. Pick one family. Read the listing like a contract. Photograph the date and write the log. Seven sentences, one per step, in the only order that works. Everything else on this page is commentary on those sentences — the worked examples, the troubleshooting, the adaptations for rentals, parents and deadlines — and commentary can be forgotten safely, because the sentences regenerate it: a chooser holding the seven sentences will reinvent the sketch, the drill and the log unaided, in the aisle, under time pressure, which is the final test any procedure should be designed to pass. Teach the sentences to the household’s other adults while you are at it; a procedure known by one person is a preference, and a procedure known by the household is a system. Write them on the log’s first page, above the entries — the summary and the record in one place, findable by anyone, forever.
Wrap-Up
Seven steps, in order: count, hazards, sensing, power, network, listing, log. Follow them and the choosing stops being a comparison of advertisements and becomes what it should be — a description of your house, answered by hardware. Start with the floor plan tonight; the rest of the procedure is downhill from there. Keep the sketch and the log together when you finish: they are the choosing, preserved — and in ten years, when the end-of-life chirps begin, they are the difference between repeating this hour and merely repeating its basket. And if the hour felt long, weigh it against the decade it governs: seven steps, once, is the cheapest insurance paperwork a household ever completes.

