Kettlewick — Playbook

At a glance

Kettlewick

Cover found before the kettle boils

Small agencies run their week on a spreadsheet and the owner is the only one who knows who can cover, but one sick text still means eight calls before 8am, so Kettlewick asks the right three carers and the first yes fills the visit.

Why this, for them

“every Monday I lose an hour to the cover call, and so does every owner I know”

It worked when: Five agencies I don't know run their Monday on it.

The problem they saw

Every Monday morning — and every Sunday night — the owner rebuilds cover by phone.

Where it goes

Every small agency in the county runs its week on it, and the phone tree is a story the owners tell new staff.

their aspiration

5 / 13canvas cells backed by graded evidence

4signals · top commitment

4features shipped · 2 not yet

The canvas, by what backs each cell

People Problem Promises Story Modes of Engagement Business Model Cost Structure What it takes to deliver Key Partnerships Metrics Risks & Harms Build or buy?Principles

commitment 1observed 2stated pain 2asserted 8

counted from the canvas, docs/evidence and the FEAT records

1 · Vision

Every small agency in the county runs its week on it, and the phone tree is a story the owners tell new staff.

In a few years

Every small agency in the county runs its week on it, and the phone tree is a story the owners tell new staff.

aspirationdocs/ideas/IDEA-001-kettlewick.md · in_a_few_years

Why this, and what "it worked" looks like

“every Monday I lose an hour to the cover call, and so does every owner I know”

Mainly about: a problem of their own

It worked when: Five agencies I don't know run their Monday on it.

asserteddocs/ideas/IDEA-001-kettlewick.md · capture log · why · motivation · success_looks_like

Principles — what we'll hold when it's costly

Never sell carer data. Nothing after 8pm. The owner is the last word on every fill.

assertedcanvas · Principles · rev. 2026-08-30

Who is building it

Marta Kowalczyk founder — Ran a twelve-carer agency for six years and did the Monday cover call every week of it.

Ola Bennett cofounder — Built the texting for a taxi firm's night desk; knows what a 7 a.m.

the team, in full →

asserteddocs/team · 2 people

The story, as text — your lines in order, for an About page or an intro email: Copy takes it

Small agencies run their week on a spreadsheet and the owner is the only one who knows who can cover, but one sick text still means eight calls before 8am, so Kettlewick asks the right three carers and the first yes fills the visit.

Why: “every Monday I lose an hour to the cover call, and so does every owner I know” It worked when: Five agencies I don't know run their Monday on it.

The problem: Every Monday morning — and every Sunday night — the owner rebuilds cover by phone. One sick text means eight calls. The visit is the thing that can't slip; the owner is the only one who knows who can take it.

How it's gone: 2026-05-20: three owners described the Monday cover call unprompted. 2026-06-14: Build the cover flow for the office screen first; the phone waits (reversed by DEC-002). 2026-07-02: The cover flow is phone-first; the office screen is the same page, wider. 2026-07-20: Kettlewick never reads a carer's location, never ranks carers, and sends nothing between 8pm and 6:30am. 2026-09-05: one owner pre-paid a month before the card form existed.

In their words: “it just texts the right three” (an owner, showing it to another owner at the school gate)

Where it goes: Every small agency in the county runs its week on it, and the phone tree is a story the owners tell new staff.

The quotes are evidence, not permission: ask before they go somewhere public.

asserteddocs/BRAND.md · the venture IDEA · canvas · docs/evidence · docs/decisions

2 · Product

Shift cover by phone for small home-care agencies — one tap when a carer calls in sick, and the cover is found before the kettle boils.

What we're testing next — the assumption it serves: owners will pay £4 a carer for cover alone, without the schedule — one has; the next four haven't been asked.

put the card form (FEAT-003) in front of the three owners who filled twice; ask for the first month, not a promise.

What would change the plan: two of three say no to paying for cover alone → the schedule question reopens (DEC-002's falsifier, early).

the founder's betIDEA-001-canvas.md · What we're testing next

How we're learning — and what was built for it

Last heard from someone outside 29 days ago.

one owner pre-paid a month before the card form existed commitment-test

What it changed: DEC-002 · Phone-first, and the schedule stays a spreadsheet · 2026-07-02
2 things heard since haven't changed anything on record

Ask someone next — /interview preps the call.

Usage: Nine owners on since 2026-08-28; six filled a visit a second Monday; one has paid. · 2026-09-07

2 met · 4 not in front of anyone yet · 2 half built

The Monday view — works
for: The owner sees what's uncovered before the first text arrives.
observed-behavior an owner covered a Monday visit from the school gate · 2026-08-30
The cover flow — works
for: An owner covers a visit from their phone in under five minutes without a call.
observed-behavior an owner covered a Monday visit from the school gate · 2026-08-30

Not in front of anyone yet · ½ = half built

Import an owner's carer list… · ½Take the first £4 · ½Cancel in one tapQuiet hours

6 built · 2 metdocs/ideas/FEAT-*.md · docs/evidence (about:)

How it works — how it shows up in someone's day — and why now

7:40 on a Monday. A text: "can't do Mrs H at 9". The owner taps the visit, Kettlewick asks the three carers who are free, qualified and close; Priya says yes at 7:43; the owner is already making tea. What changed: the schedule went to phones in 2024–25, the cover call never did. Smallest workflow: the ask, the yes, the fill.

EVID ×1 · observed-behaviorcanvas · Story · rev. 2026-08-30

What it looks like — 2 screens

docs/product/screens

What it is today — the current shape, in the founder's words

The best articulation so far.

What: One tap when a carer calls in sick. The schedule stays a spreadsheet; Kettlewick only does the cover.
Who it's for: owner-operators with 3 to 15 carers who do the Monday schedule by hand.
Why now: owners moved the schedule to their phones in 2024–25; the cover call didn't follow.
Smallest version that proves it: a text goes out to the three carers who can take the visit; the first yes fills it; the owner sees it filled.

asserteddocs/ideas/IDEA-001-kettlewick.md · ## Current shape · since 2026-05-12

What it is not

agency software. A marketplace. A scheduling tool.

asserteddocs/BRAND.md · What it is NOT

3 · Customers

52, owns a nine-carer agency in a market town; runs it from the kitchen table before the school run, and from the school gate after it.

Marta — primary

52, owns a nine-carer agency in a market town; runs it from the kitchen table before the school run, and from the school gate after it.

Sunday 8:40pm or Monday 7:40am: a text says "can't do tomorrow" and the phone tree starts — eight calls, an hour, a spreadsheet open on the laptop and a biro on the back of an envelope.

the full card, in Design →

synthetic 40% · real 60%docs/personas/marta.md

Priya — secondary

34, a carer with two agencies, a car, and a school run of her own; eleven years in care, four with Marta.

7:41am, between dropping her son and her first visit, a text asks if she can take Mrs H at 9; she's already dressed, and she answers with one thumb at a red light.

the full card, in Design →

synthetic 100% · real 0%docs/personas/priya.md

4 · Problem

Every Monday morning — and every Sunday night — the owner rebuilds cover by phone.

The problem

Every Monday morning — and every Sunday night — the owner rebuilds cover by phone. One sick text means eight calls. The visit is the thing that can't slip; the owner is the only one who knows who can take it.

EVID ×2 · stated-paincanvas · Problem · rev. 2026-08-30

How it shows up, and why now

In someone's day, with what changed that makes it newly possible: How it works, in Product →

canvas · Story

5 · Market

Owner-operators of small home-care agencies — 3 to 15 carers, run from a kitchen table or a back office.

How many, and how do you know — the People cell

Owner-operators of small home-care agencies — 3 to 15 carers, run from a kitchen table or a back office. About 6,400 such agencies in England (register extract, 2026-06-02, docs/source). We know from nine of them, in person.

EVID ×1 · stated-paincanvas · People · rev. 2026-08-30

Research you've imported — 2 items in docs/source/

2026-07-15-owner-survey-notes.md 2026-07-15
2026-06-02-register-extract.csv 2026-06-02

imported · not read heredocs/source/ · /import

6 · Competition

3 on the field; 2 of them key.

Who else fixes it — docs/competition/README.md

Rival	What it is	Pricing	Why they might win	Where they're weak	Checked
Shiftwise	direct — agency scheduling software with a cover module	£12 / seat / month	every inspector already recognises their export; owners have it	office-first; the cover module needs the office screen	2026-09-02
CareSheet	direct — a shared spreadsheet template with a texting add-on	£0 + texts	free, and it's the spreadsheet owners already have	the add-on texts everyone, not the right three	2026-09-02
The phone tree	in evidence — doing nothing, by hand	an hour a Monday	it works, it's known, it costs no money	it costs the owner's hour and the carer's evening	2026-09-02

3 on the field · 2 key · 0 staledocs/competition/README.md · updated 2026-09-02

Key rivals

direct, or named by a real person in evidence

Shiftwise — key

They might win because every inspector already recognises their export; owners have it

Where it breaks:

owners rate the phone app poorly — "it logs me out on every visit"
cover is still a phone call: the module lists who's free, then you ring them
the office screen is where it works; Monday morning happens on a phone

checked 2026-09-02docs/competition/shiftwise.md

CareSheet — key

They might win because free, and it's the spreadsheet owners already have

Where it breaks:

the add-on texts every carer, every time — owners turn it off after the first complaint
no yes: the owner still rings to confirm who's taking it

checked 2026-09-02docs/competition/caresheet.md

Also on the field

watch — real, filed, not yet in anyone's mouth · one line each

Watch list

The phone tree — it works, it's known, it costs no money

1 on watchdocs/competition/README.md · watch rows

7 · Canvas

Switch the frame — Humane, Lean, BMC — and the boxes move; the words don't. A dashed box is a question nobody has answered. Two cells stay on the page in every frame.

Humane Product Canvas · Ajesh Shah

1 · Human foundation

who you serve, the tension they carry, the value you promise

PeopleCustomer segmentsCustomer segments

Who are you designing for & what matters to them? Who exactly has the painful problem — how many, and how do you know?

Owner-operators of small home-care agencies — 3 to 15 carers, run from a kitchen table or a back office. About 6,400 such agencies in England (register extract, 2026-06-02, docs/source). We know from nine of them, in person.

EVID ×1 · stated-paincanvas · People · rev. 2026-08-30

ProblemProblem

What real human tension are you solving? Is it urgent, frequent, expensive, or emotionally painful?

Every Monday morning — and every Sunday night — the owner rebuilds cover by phone. One sick text means eight calls. The visit is the thing that can't slip; the owner is the only one who knows who can take it.

EVID ×2 · stated-paincanvas · Problem · rev. 2026-08-30

PromisesUnique value propositionValue propositions

What emotional / relational value will it deliver? The sharp promise: "We help X do Y without Z."

We find cover before the kettle boils. One tap; the right three carers get the ask; the first yes fills the visit; the owner sees it filled.

EVID ×1 · observed-behaviorcanvas · Promises · rev. 2026-08-30

2 · Product expression

how it shows up in a life, how people engage, how it sustains itself

StorySolution

How does your product show up in someone's life? What changed that makes this newly possible?

7:40 on a Monday. A text: "can't do Mrs H at 9". The owner taps the visit, Kettlewick asks the three carers who are free, qualified and close; Priya says yes at 7:43; the owner is already making tea. What changed: the schedule went to phones in 2024–25, the cover call never did. Smallest workflow: the ask, the yes, the fill.

EVID ×1 · observed-behaviorcanvas · Story · rev. 2026-08-30

Modes of EngagementUnfair advantageCustomer relationships

How do people interact with your product in a humane way? Does it respect time, attention, agency?

Self-serve, phone-first, and quiet: nothing pings after 8pm, no location tracking, no league table of who says yes. The owner's own carers are the only people asked. Why us: Marta ran a twelve-carer agency for six years and did this call every week; the product is her Monday, not a demo of one.

assertedcanvas · Modes of Engagement · rev. 2026-08-30

Business ModelRevenue streamsRevenue streams

How will you sustain this without compromising your promise? Who pays, how much — and how do the first 100 find you?

£4 per carer per month, flat, paid by the agency; free for the carers. The first 100 find us through the county owners' group and the register (owners talk to owners); usage compounds because every fill is a text a carer keeps. Channels that keep working after the first 100: the owners' group, and the carers who move agencies.

EVID ×1 · commitmentcanvas · Business Model · rev. 2026-08-30

Cost StructureCost structureCost structure

What does it actually cost to serve one person for a month?

About 40p per carer per month at nine agencies — texts and hosting; no model in the loop. Fixed: one laptop, one phone number. Runway: Ola part-time on savings until December; Marta full-time on the agency sale. If cover isn't earning by then, it's a hobby.

assertedcanvas · Cost Structure · rev. 2026-08-30

What it takes to deliverKey activities · Key resources

What do you need to have and do to keep the promise?

A phone number that sends texts reliably at 7 a.m.; the owner's schedule as a CSV every week; Ola's evenings. Nothing else is load-bearing.

assertedcanvas · What it takes to deliver · rev. 2026-08-30

Key PartnershipsKey partnerships

Is there anyone whose cooperation this can't work without?

None load-bearing. The register is open data; the text provider is swappable in a day.

assertedcanvas · Key Partnerships · rev. 2026-08-30

Channels

Lean and BMC have a box for this; the humane canvas answers it inside Business Model — how the first 100 find you. One answer, never rendered twice.

canvas · Business Model

3 · Stewardship

impact, risks, and the values that guide decisions

MetricsKey metrics

What does meaningful success look like — for people and planet? Real pull: activation, retention.

It worked = five agencies we don't know run their Monday on it. Real pull: a visit filled by Kettlewick without the owner's phone (activation), and the same owner filling again the second Monday (retention). Nine owners on; six filled twice.

assertedcanvas · Metrics · rev. 2026-08-30

Build or buy?

Is the honest move to build this — or would using / buying something existing get the same outcome?

Build the cover flow only. Buy the texting; keep the schedule where it is (the spreadsheet). Shiftwise does schedules; we don't.

assertedcanvas · Build or buy? · rev. 2026-08-30

Two questions this canvas asks that Lean and BMC don't

they render in every frame — the floor, not a footnote

Risks & Harms

What could unintentionally go wrong? Who might be harmed or excluded?

A visit marked covered that isn't — a person doesn't get their care. Carers pressured to say yes at 7 a.m. — refused by design (no reply is fine, nothing shows who declined). Location tracking — refused. The regulator names a duty of care to the person visited; our stance: the owner confirms every fill, Kettlewick never marks a visit covered on its own.

assertedcanvas · Risks & Harms · rev. 2026-08-30

Principles

What values will guide your decisions? The non-negotiables you'll hold even when it's costly.

Never sell carer data. Nothing after 8pm. The owner is the last word on every fill.

assertedcanvas · Principles · rev. 2026-08-30

8 · Business model

£4 per carer per month, flat, paid by the agency; free for the carers.

Who pays, how much

£4 per carer per month, flat, paid by the agency; free for the carers. The first 100 find us through the county owners' group and the register (owners talk to owners); usage compounds because every fill is a text a carer keeps. Channels that keep working after the first 100: the owners' group, and the carers who move agencies.

EVID ×1 · commitmentcanvas · Business Model · rev. 2026-08-30

What it costs to serve

About 40p per carer per month at nine agencies — texts and hosting; no model in the loop. Fixed: one laptop, one phone number. Runway: Ola part-time on savings until December; Marta full-time on the agency sale. If cover isn't earning by then, it's a hobby.

assertedcanvas · Cost Structure · rev. 2026-08-30

Prior capital and ownership

none — sole owner, no money in but my own

asserteddocs/ideas/IDEA-001-kettlewick.md · prior_capital

The ask — the capital mentor's read

Not yet — one agency has paid and eight haven't; the raise question opens when five agencies we don't know pay a second month. Until then the model question is the only live one: does £4 a carer for cover alone hold, or does it need the schedule?

docs/dossier/business-2026-09-05.md · 2026-09-05

docs/dossier/mentor-capital.md

9 · Evidence

one owner pre-paid a month before the card form existed

How sure, over time — the strongest grade on record by date; decisions on the axis

4 signals from 2026-05-20 to 2026-09-05; the strongest so far is commitment (EVID-004, 2026-09-05); 3 decisions on the axis.

from evidencedocs/evidence · docs/decisions · dates and grades only

The ladder — 4 signals, graded — the grade is the founder's, the count is the file's

commitment1

observed-behavior1

stated-pain2

Grade	Date	Signal	Method	Bears on	Record
commitment	2026-09-05	one owner pre-paid a month before the card form existed	commitment-test	owners will pay £4 a carer for cover alone	EVID-004
observed-behavior	2026-08-30	an owner covered a Monday visit from the school gate	observation	owners will cover a visit from their phone	EVID-003
stated-pain	2026-06-04	an owner said the hour is the job she hates most	interview	owners lose Monday to the cover call	EVID-002
stated-pain	2026-05-20	three owners described the Monday cover call unprompted	interview	owners lose Monday to the cover call	EVID-001

docs/evidence · 4 records · bodies stay in the files

10 · Learnings

the card form works on my phone; the invoice line names the carers.

How it grew — the turns, each in the record's own words

Once upon a time2026-05-12The idea

Shift cover by phone for small home-care agencies — one tap when a carer calls in sick, and the cover is found before the kettle boils.

IDEA-001

But one day2026-05-20The first signal

three owners described the Monday cover call unprompted

EVID-001 · stated-pain

Because of that2026-06-14The first bet

Build the cover flow for the office screen first; the phone waits.

DEC-001 · superseded

But then2026-07-02What changed

The cover flow is phone-first; the office screen is the same page, wider.

DEC-002

Until finally2026-09-05Someone committed

one owner pre-paid a month before the card form existed

EVID-004 · commitment

And one dayin a few yearsWhere it goes

Every small agency in the county runs its week on it, and the phone tree is a story the owners tell new staff.

their aspiration

the venture IDEA · docs/evidence · docs/decisions

2026-09-08

Landed: the card form works on my phone; the invoice line names the carers.

Surprises / decisions: an owner asked for a register import before paying — FEAT-004 captured as ready.

docs/devlog.md · 2026-09-08

2026-09-05

Landed: one owner pre-paid a month by bank transfer before the form existed (EVID-004).

docs/devlog.md · 2026-09-05

2026-08-30

Landed: watched an owner cover a visit from the school gate, no call (EVID-003). Enriched Marta's persona from it.

Surprises / decisions: the office screen was the wrong bet — DEC-001 marked wrong, DEC-002 stands.

docs/devlog.md · 2026-08-30

2026-08-28

Landed: shipped to nine owners.

docs/devlog.md · 2026-08-28

2026-08-28 · IDEA-001

shipped to the first nine owners. Marta covered a Monday visit from the school gate.

docs/ideas/IDEA-001-kettlewick.md · 2026-08-28

2026-07-15 · IDEA-001

owner survey notes: the phone tree is the job they hate most; the spreadsheet they'd keep.

docs/ideas/IDEA-001-kettlewick.md · 2026-07-15

2026-07-02

Landed: the Monday view on a phone.

Surprises / decisions: DEC-002 — phone-first; the schedule stays a spreadsheet.

docs/devlog.md · 2026-07-02

2026-06-04

Landed: three calls; the persona has a real share for the first time.

Surprises / decisions: nobody wants a schedule. They want the cover done.

docs/devlog.md · 2026-06-04

4 earlier entries in docs/devlog.md · docs/ideas/IDEA-001-kettlewick.md

11 · Decisions

Kettlewick never reads a carer's location, never ranks carers, and sends nothing between 8pm and 6:30am.

No location tracking, no league table, nothing after 8pm — DEC-003

Kettlewick never reads a carer's location, never ranks carers, and sends nothing between 8pm and 6:30am.

Falsifier — If owners leave for a rival because of tracking, by July 2027, the cost was real. by 2027-07-20

one-wayfounderdocs/decisions/DEC-003-no-tracking.md · 2026-07-20

Phone-first, and the schedule stays a spreadsheet — DEC-002

The cover flow is phone-first; the office screen is the same page, wider. The schedule stays the owner's spreadsheet; Kettlewick reads it and never edits it.

Falsifier — If half the fills by December start from the office screen, the phone bet was wrong. by 2026-12-01

costlyfounderdocs/decisions/DEC-002-phone-first.md · 2026-07-02

Office-first cover — DEC-001

Build the cover flow for the office screen first; the phone waits.

Falsifier — If three owners cover from a phone before August, this was wrong. by 2026-08-15

superseded by DEC-002

reversiblefounderoutcome recordeddocs/decisions/DEC-001-office-first.md · 2026-06-14

Build decisions — 1, one line each

The brand anchor: kettle copper, brown-black, two radii, one shadow DEC-004 · 2026-09-02

docs/decisions · scope: build

12 · Risks & harms

A visit marked covered that isn't — a person doesn't get their care.

Who could this harm, and how

A visit marked covered that isn't — a person doesn't get their care. Carers pressured to say yes at 7 a.m. — refused by design (no reply is fine, nothing shows who declined). Location tracking — refused. The regulator names a duty of care to the person visited; our stance: the owner confirms every fill, Kettlewick never marks a visit covered on its own.

assertedcanvas · Risks & Harms · rev. 2026-08-30

Trust — what you collect, in plain terms

We hold the schedule you give us and the phone numbers on it; nothing else. Carers' locations are never read. Nothing is sold, shared or used to train anything. Delete your agency and it is gone in a day, texts included. To ask about your data, text the number you already have.

asserteddocs/trust/TRUST.md · 2026-10-04

13 · Health

Nine owners on since 2026-08-28; six filled a visit a second Monday; one has paid.

The health read — 2026-09-07

Nine owners on since 2026-08-28; six filled a visit a second Monday; one has paid. Pre-fit, honestly — n is under ten, so this is a read, not a curve, and the next move is to watch the three who didn't fill twice.

a verdict, dateddocs/health/HEALTH-2026-09-07.md

What is measured — 2026-08-30

One activation metric: a visit filled by Kettlewick without the owner's phone. One retention curve: the same owner filling again the second Monday. Six events, no more: ask sent, yes received, fill confirmed, fill cancelled, owner opened Monday view, carer opened ask. No location, no ranking, aggregate by default.

asserteddocs/measure/MEASURE-2026-08-30.md

14 · Team

Ran a twelve-carer agency for six years and did the Monday cover call every week of it.

Marta Kowalczyk — founder · @marta

Ran a twelve-carer agency for six years and did the Monday cover call every week of it.

Brings: the owners' trust; the schedule in her head; the inspector's questions by heart.
Doesn't: code. Ola does.

Marta owned a home-care agency until 2025 and sold it to build the thing she wished she'd had. She still covers a visit herself when a friend's agency is short.

docs/team/marta.md

Ola Bennett — cofounder · @ola

Built the texting for a taxi firm's night desk; knows what a 7 a.m. message has to survive.

Brings: the build, evenings and weekends until December; a horror of anything that pings at night.
Doesn't: the care domain — Marta translates.

Ola has shipped small things for small firms for a decade and never wanted a big one.

docs/team/ola.md

Who is missing — 1 role, named

A designer, part-time — Someone who has drawn a screen a carer reads at 7 a.m. with one thumb and a child on the other arm.

Brings: the carer side, which is plainer than it should be (see How we build).
Doesn't: need to be full-time — two days a month from November.

asserteddocs/team/missing-designer.md

15 · Brand

it will not rank carers, track carers, or text anyone after 8pm.

Current shape — nascent

Who it's for

owner-operators who do the Monday schedule by hand and would keep the spreadsheet forever.

What it promises

cover found before the kettle boils.

What it refuses

it will not rank carers, track carers, or text anyone after 8pm. DEC-003

How it sounds

plain, not clever (costs us delight); calm, not urgent (costs us the "act now" conversion); the owner's words, not ours.

Example "Mrs H at 9 is covered. Priya said yes." (the confirm text owners forward to each other) EVID-003

What it is NOT

agency software. A marketplace. A scheduling tool.

What they use instead today

the phone tree, an hour of calls every Monday; a few try CareSheet's add-on, which texts everyone. EVID-001

The name, and why

the kettle is the owner's Monday morning; the wick is the bit that catches. It means nothing to a stranger and everything to an owner.

7 of 7 knowndocs/BRAND.md · 2026-09-02

Words that landed — theirs, quoted; credited by role, dated, traced to the record

“it just texts the right three”

an owner, showing it to another owner at the school gate · 2026-08-30 · EVID-003

the right three is the promise people repeat

“the thing that finds cover”

an owner, nine-carer agency, called it · 2026-06-04 · EVID-002

the verb is the brand; the product name isn't

from evidencedocs/BRAND.md · What we've learned · 2 rows

Origin, as it happened

2026-05-12 — another Monday lost to the cover call: a text at 7:40, eight calls, one yes. I wrote the idea down that morning. I want the hour back, and so does every owner I know.
2026-05-20 — three owners, three calls, the same hour described unprompted. Nobody asked for a new schedule; they asked for the cover to be done. — EVID-001

asserteddocs/BRAND.md · Origin

The anchor — what is decided

#B84E12 accent

Cover found before the kettle boils

docs/BRAND.md · frontmatter · tokens and type, in Design →

How it changed — each change with its why, newest first

2026-07-15 voice: the owner's words over ours (was: formal and complete) — official wording read as being told off — an owner said the first reminder text "sounded like the council"

asserteddocs/BRAND.md · What we've learned · shape: / voice: rows

16 · Values

Owners first

Principles — the short form

Never sell carer data. Nothing after 8pm. The owner is the last word on every fill.

assertedcanvas · Principles · rev. 2026-08-30

Owners first

every screen is designed at the owner's desk on a Monday morning, not in a demo.

costs the carer side is plainer than it could be.

docs/BRAND.md · How we build

No league tables

we never rank people, even when an owner asks.

costs some owners ask, and leave.

docs/BRAND.md · How we build

Nothing after 8pm

the product is asleep when the carers are.

costs a late cancellation waits until 6:30.

docs/BRAND.md · How we build