THE CREATIVE BRIDGE · BUILD SPEC
AUGUST 2026

Build spec

The design funnel

A café owner answers eight either/or questions on your website and watches their own ordering app take shape as they tap. Thirteen taps worst case, eleven on a smooth run. This spec covers the question sequence, the progress and celebration mechanics, and where an AI model does and doesn't belong.

The two headline findings

The funnel needs no AI at runtime. Eight binary questions produce an 8-bit answer — that's a lookup table, not an inference problem. A model in that path buys non-determinism, a network dependency and a failure mode in exchange for nothing, and costs you the ability to answer "why did my app come out green?" A model earns its keep after the last tap: parsing their menu, and writing the copy in their voice.

Progress bars correlate with lower completion in the wild — the largest study here (25,080 real surveys) found exactly that. It doesn't mean drop the countdown you asked for; it means build it the specific way that works — discrete steps, endowed with a head start, counting down — and measure it rather than trusting it.

Architecture

Where the model is, and isn't

THE FUNNEL — PLAIN CODE, NO MODEL 8 taps lookup table answers → tokens live preview re-renders each tap AFTER THE REVEAL — WHERE A MODEL EARNS ITS COST Parse their menu photos → items, sizes, modifiers Write their copy in the voice they picked Ongoing: review replies, campaign drafts WHY IT STAYS DETERMINISTIC Same answers → same design, every time Testable, instant, works offline, free You can always answer "why did it come out this way?"
Two model calls per café, not one per tap. The funnel itself is a lookup table — which is why it can be instant, free, and identical every time. Variety comes from more axes in the table (palette × type × density × imagery multiplies into thousands of combinations), not from inference.

Which model

Claude Sonnet 5, not Opus 5 — and the reason isn't benchmarks. At this volume the entire monthly spread between the cheapest and most expensive credible option is under A$60, so this is not a cost decision. It's an integration decision: you already build with Claude, and one provider means one key, one SDK, one billing line, one place to debug. Opus 5 is real money for no gain on copywriting; Haiku's smaller context is the one thing in the lineup that could choke on a long PDF menu.

One exception worth testing. Menu parsing is the highest-value model task — it replaces 45–90 minutes of typing per client, and it's the difference between a A$1,000 build being profitable. It's also the one place the evidence points away from Anthropic: Google's models lead document-parsing benchmarks at roughly a third the cost. But none of those benchmarks test café menus, where the hard part isn't reading the text — it's interpreting size tiers, modifier groups and "+50¢ oat". Run a bake-off on 20 real messy menus (angled phone photos, a chalkboard, a two-column PDF, one handwritten), scored on items, prices and modifiers correct. If Gemini wins clearly, use it for menu parsing only — it's one bounded function.

Don't self-host (break-even is ~400× your volume), don't adopt a model gateway (~100× your spend), and never run client menus through a free tier that may train on them.

The sequence

Eight questions, ranked by what each one decides

Six show two rendered phone screens; two are text, because tone doesn't photograph. Every question varies exactly one attribute and holds everything else identical — the most common failure in visual quizzes is a pair that differs in three ways at once, so the tap tells you nothing about which one drove it.

Order runs easiest and most enjoyable first, most committing last. The largest study of survey completion found that the first question's type predicts whether people finish, so the opener is a pure image tap — never a text field, never a name.

START · TAP 1"8 to go"

Design my café's app

Phone mockup already on screen showing a neutral ordering page. Progress shows 10 dots with 2 already filled.
The head start is the single best-evidenced trick here — a car-wash study of 300 customers found a 10-stamp card with 2 pre-filled beat an 8-stamp empty card 34% to 19% for identical actual work. Preload every image here; a spinner mid-quiz recreates the exact failure the progress research warns about.
Q1 · MOOD · TAP 2"7 to go"

Which feels more like your café?

Same ordering screen twice — warm, rustic, handmade vs crisp, modern, minimal.
🎉 Celebration 1 (inline, zero taps): the mockup re-renders in their direction. "Warm and rustic. That's most of the hard part decided." Governs everything downstream, and rewards the very first commitment by proving the machine is real.
Q2 · COLOUR · TAP 3"6 to go"

Which colours feel right?

Identical layout, palette only — terracotta/cream/brown vs sage/slate/off-white.
Q3 · LIGHT OR DARK · TAP 4"5 to go"

Light and airy, or dark and cosy?

Identical design, background only.
Commercially real rather than cosmetic — a menu read in a dim café behaves differently from one read on a sunny footpath.
Q4 · IMAGERY · TAP 5 · MIDPOINT"4 to go"

Big photos of the food, or clean type and no photos?

The highest-value question on the sheet.
🎉 Celebration 2 — the biggest visible transformation, placed at the midpoint where abandonment peaks. "Look at that. It's starting to look like yours." This question also decides whether the café needs a photoshoot, and most cafés have poor photography — one tap that can save you a project's worth of grief.
Q5 · MENU LAYOUT · TAP 6"3 to go"

Show a few things beautifully, or lots of things quickly?

Big photo cards vs compact list.
Quietly qualifies menu size before you've quoted the build.
Q6 · LETTERING · TAP 7"2 to go"

Which lettering?

Two rendered menu headers, same words — characterful serif vs geometric sans.
Q7 · TONE · TAP 8 · TEXT"1 to go"

Which sounds like you?

"Morning! What are we having?" vs "Order — ready in 5 minutes."
Real strings from the app, not descriptions of a tone. This answer feeds the copy the model writes later.
Q8 · PRIORITY · TAP 9 · TEXT"Done"

What matters more right now — moving the queue, or bringing people back?

Decides whether the loyalty card sits on the home screen.
🎉 Celebration 3 — the only confetti in the whole funnel. "That's your café's app designed. Eight taps." Most committing and most flattering question, so it goes last.
REVEAL + DETAILS · TAPS 10–13all 10 filled

Their finished home screen, full size, unblurred

Then beneath it: "Want it live? Tell us where to send the working link." Café name, name, email required; phone optional.
Show the design first, then ask — the reveal is the incentive, so spending it before the ask turns you into a tollgate. This screen sits outside the count and says so, because the countdown must never lie.

Mechanics

Progress, celebration, and the tap budget

The finding that shapes the design

Across 25,080 real web surveys, the ones without progress bars had higher completion than the ones with. A separate controlled study found the mechanism: it's not the presence of progress, it's the rate — early feedback showing slow progress increases abandonment, while intermittent feedback tested best of all.

So the countdown you asked for stays, but built the specific way that survives this: discrete dots, not a creeping percentage (at 10 steps each tap is a visible chunk; a bar crawling 0→10% is precisely the failure mode), endowed with 2 filled so question one lands at 30% rather than 10%, and counting down because "5 to go" is concrete where "50% complete" is abstract. Then instrument it and A/B it against a plain "3 of 8" — the evidence is strong enough to design around but not strong enough to trust blind.

Celebrations that don't patronise

Three, all inline and zero-tap. None gets its own screen or a Continue button — three interstitials would eat 20% of the tap budget outright.

Every one says what the answer did: "Warm and rustic — that's most of the hard part decided." Never "Great job!" — congratulating a business owner for tapping a picture reads as contempt.

The best celebration isn't a celebration: the mockup visibly re-rendering in their choices after every tap. Intrinsic payoff, no praise, no tap.

The tap budget

1 start · 8 questions · 4 for details and submit = 13 worst case

Eleven on a smooth run, because the phone keyboard's Next key moves between fields without taps. That leaves two in reserve for a back-tap and the result-page call to action.

Three rules hold it together: auto-advance on tap (no Next button on question screens), zero-tap celebrations, and one details screen — never split contact fields across pages.

Mobile and accessibility

Both option cards stack full-width in the lower two-thirds of the screen — roughly half of phone use is one-handed with a thumb, and the top corners are the worst real estate available. Never side by side on a phone: three-across or side-by-side shrinks the images, and shrunken images defeat a visual-preference question entirely. No scrolling on any question screen. Auto-advance with a 350–450ms transition, long enough to see the mockup update and short enough not to drag.

Alt text describes the difference, not the position — "Ordering screen in warm terracotta and cream", never "Option A". Announce progress in a polite live region ("Answered. Five questions to go.") and move focus to the new question heading on auto-advance, or keyboard and screen-reader users are stranded mid-funnel. Wrap the confetti in prefers-reduced-motion: no-preference and swap an instant static state change underneath — and note the popular confetti library animates regardless unless you gate it yourself.

Give every pair a low-emphasis "not fussed" link. Forced binaries with no escape produce irritation and random data, and one tap either way keeps the budget intact.

What kills it

The failure modes, ranked

MistakeWhy it's fatal
Celebration screens with a Continue buttonThree of them consume a fifth of the entire tap budget. Inline only, always.
Pairs that differ in three ways at onceIf A is dark and serif and photo-led, the tap tells you nothing. Vary one attribute; hold the rest identical. Otherwise the "design" is theatre.
Asking about taste in words"Modern or rustic?" — nearly everyone says modern, and the words map to different pictures in different heads. Show rendered screens.
Over-promising the countIf it says "3 to go" it must be exactly 3. Under-promising is safe; over-promising is a documented abandonment driver.
A canned resultEight questions must not collapse into one of four templates. If the reveal isn't visibly bespoke, the funnel is exposed as a lead-capture costume — and your credibility with a small-business owner is the actual product.
Pairs too similar to tell apartIf the difference isn't visible in about a second, on a phone, at arm's length, the answer is noise. Exaggerate the contrast beyond what you'd ship.
Email gate before the revealYou lose people at the worst possible moment — after they've done all the work. Show the design, then ask where to send the working link.
Required phone numberThe highest-friction field on any form. This audience will give it to you on the call anyway.
Stock photos of cafésAbstract and lifestyle imagery measurably fails at this job. Show the product — a phone mockup of the actual ordering screen.
Assets loading mid-quizA spinner at question four is the slow-progress failure mode by another route. Preload everything on the start screen.

Decisions still open

Three calls to make before building

  1. Binary everywhere, or three options where three are natural? The honest evidence is that drop-off scales with the number of questions, not options per question — and a third option costs the same single tap while returning more information. Binary is defensible here on layout grounds (two stacked cards fit a phone; three don't) and on brand grounds. Just don't contort a genuinely three-way question into two binaries: that spends two taps to learn less.
  2. Does the price go on the result page? Publishing "A$1,000 setup, A$99/month" beneath the reveal costs a few leads and saves your son hours of unqualified calls. At eight taps of self-selection you're already getting few, good leads rather than a big list — I'd publish it.
  3. Who does the funnel actually run for? Cafés who've met you, as a follow-up that turns a conversation into a real design — or cold traffic from the website. It changes the copy at the top and nothing else structurally, but it changes what "good completion" means.

Prepared August 2026 for The Creative Bridge (Adelaide, SA). Evidence grades in the underlying research separate peer-reviewed findings from vendor marketing; where a claim is vendor-sourced it is marked as directional in the source notes. Companion documents: the agent-run research brief, and the live demo at demo.lotushosting.org.