Case studies · our own work

Our own work, before and after.

Our own designs, made easier to use. The painting business and its figures are fictional.

Earlier painting example on a phone: headline and six source tiles; the business view starts at the bottom of the screen.
Before · 9 NovSix source tiles first. The output starts below the screen.
Rebuilt painting example on a phone: headline, then sample quote Q-2291 for $44,104 CAD with line items and a Needs your review label.
After · 10 NovA sample quote, with labour and cost per line, on the first screen.
Case study of our design work · the contractor, job and figures are fictional

Turning a painting-business pitch into a quote you can try.

Our earlier example explained where information comes from. The new one starts with the job a painting contractor actually prices, then follows it through to the crew and the next bid.

What we started with

Sources first, output later

Six tiles for email, bid boards, drawings, takeoffs, rates and crew. On a phone, the result card began at the bottom of the screen.

What we changed

The quote comes first

Line items by surface, count and length. One action: see what an addendum changes. Then review, award and crew handoff.

What you can use

A working sample job

Change a rate and the quote, review and crew plan update together. Invalid rates are stopped, not totalled.

Painting example after Addendum 02: walls 18,400 to 21,600 sq ft, frames 64 to 76, new epoxy line, quote version 2 at $51,746 CAD, needs review.
After the addendum: changed rows edged in gold, version 2, review reopened.

What the checks show

  • Earlier sample: 37 of 37 automated browser checks passed, in the original run and again in a rerun on 10 Nov.
  • No sideways scrolling at 320, 390, 768 and 1440 px wide, before and after the addendum.
  • The default total matches an independent recalculation ($44,104 CAD). The addendum moves it to $51,746.
  • Blank, negative, zero or non-numeric rates stop progress and show “—”.
  • axe reported no accessibility violations in the default or fully progressed state.
  • The sample runs locally; it does not read documents, send messages or connect accounts.

A scripted sample: no live document reading, integrations or sending. It is not a contractor's result, and the rates are not our prices.

Try the sample job

Open it full screen, load it here, or download the single file to keep.

Case study of our engineering work · mindsonfire.ca, 10 Nov 2026

Checking our own website the way visitors use it.

We checked the homepage on a phone and desktop, reduced the first-visit cookie banner, and kept the useful output and consent choices intact.

What we started with

A crowded first visit

The phone cookie banner took up more than half a short screen.

What we changed

Small, specific fixes

A shorter banner with expandable privacy details, and clearer lead and contact wording.

What stayed the same

The approved hero

The headline, design and consent logic stayed unchanged. The sample still responds to your business name.

Before and after the fixes

CheckBeforeAfter
Phone banner height394 px158 px
Desktop banner height191 px114 px
Phone / desktop sideways scrollingNoneNone

Lab and browser checks, not visitor data. They do not measure rankings, bookings or revenue. Checked 10 November 2026 at 390 × 740 and 1440 × 900. This is a bounded homepage check, not a claim that every page is perfect.

Idea builder before: placeholder text under Your goal in light grey.
Before · placeholder contrast 4.0 : 1
Idea builder after: the same placeholder text in a darker grey.
After · 5.4 : 1 (WCAG AA asks for 4.5)

Earlier idea-builder contrast repair shown above; homepage measurements from Playwright (Chromium).

Want the same for your business?

Tell us what you are working on. We will pick a first piece together, build it, and check it the same way.