Zagreb

Healthy Gut · what we made for you, unasked

Beforethe brief.

A look at Healthy Gut’s identity and pages, made before any brief.

We made it unasked, so you can judge the work itself before we talk. Read it in this order; each part opens on its own:

  • Your identity, sharpened: your mark, your navy and green
  • What we would do for product photography
  • One product, HoloZyme, on three landing pages
  • Tributyrin-X: the ad, then story, quiz and offer, and how we would test them
The identity upgrade, first screen on a phone Landing page A, Voices, first screen on a phone Landing page B, The plate, first screen on a phone Landing page C, Find your dose, first screen on a phone

The four parts, first screen on a phone: the identity upgrade, then landing pages A, B and C. Captured 08.11.2026.

The parts:

Identity.
01: the upgrade, our suggestion, with product photography: what we would do.
Voices.
02: landing page A for HoloZyme, led by what customers say.
The plate.
03: landing page B for HoloZyme, led by the foods people avoid.
Find your dose.
04: landing page C for HoloZyme, led by the dose for each meal.

Where the chain starts:

A woman tears bread over a plate of roasted garlic and aged cheese at her table, under a bare plaster wall. AI-generated, fictional person.Bloating? Loose stools? Foods you react to?Five questions sort out where your gut should start.Take the 5-question quiz
A woman smiles as she lifts a piece of bread at her supper table. AI-generated, fictional person.Bloating? Loose stools? Foods you react to?Five questions sort out where your gut should start.Take the 5-question quiz
A woman at a long table holds bread above her plate of roasted garlic and cheese, beside a window. AI-generated, fictional person.Bloating? Loose stools? Foods you react to?Five questions sort out where your gut should start.Take the 5-question quiz

The ad says what the page says: its headline is the story page’s headline, and its button is the story page’s button. From there the chain runs story, quiz, offer, in the same words.

Your Tributyrin-X pages, and ours:

Your two campaign pages already do real work: an opening question the reader answers in her head, practitioners quoted with their credentials, and the full review spread on the offer page. Here is what we would change, and why.

01
One path, not separate pages.
Yours: two campaign pages that do not lead into each other. The quiz page sends buyers on to a third page; the offer page to a checkout on the page itself.
Ours: one chain of three pages. The story leads to the quiz, the quiz to the offer, and the offer opens with the reader’s own answer: her concern and the plan the quiz chose for her. A reader who is ready can skip straight to the offer.
Why: the page should continue the ad’s promise in the same words (checklist C1‑01), and a story page before the product page is a recognised format for cold Meta traffic (C1‑11).
02
A quiz that ends in a plan.
Yours: one question, answered in the next line.
Ours: five short questions: what bothers her most, for how long, what she has tried, how her gut takes new supplements, and whether she has taken Tributyrin-X before. The result names her concern and picks one of your two subscription plans by your own plan notes: a first start gets Start & save, “Best for dialing in your dose”; a returning buyer gets the 3-month plan, “For once your dose is established”. It shows your dosing help, a buyer whose words match her concern, and the button. No email asked, and no dose suggested.
Why: a quiz works when its result screen carries the recommendation and the buy button, not only an answer (C2‑04). With one product there is no product to choose (C2‑07), so our quiz chooses the plan, the real decision on your page.
03
One set of numbers.
Yours: each page carries its own figures, so a buyer who opens two of them sees different numbers: four review counts (over 450, 452, 500+ five-star and 680) where your product page says 788, two customer counts (30,000+ and 40,000+), and two discounts (15% and 30%).
Ours: one set of numbers on every page, taken from your own product page: 788 reviews, with the full spread of 650 five-star, 104 four-star, 20 three-star, 7 two-star and 7 one-star. The quiz and the offer read them from one shared list, so they cannot drift apart.
Why: the rating and review count next to the product name answer the cold visitor’s first question (C1‑05), and a real spread is trusted more than a wall of five stars (C4‑01).
04
Claims you can stand behind.
Yours: some lines go further than the evidence on the page, like results “in days” or “for good”, or weight loss.
Ours: only claims your product page already makes as plain product facts: 99% pure research-grade oil, slow-release butyrate that coats the intestines, 500 mg per gel cap, free from gluten, dairy, soy, added sugar and GMOs, and the 60-day guarantee.
Why: under the FTC’s guidance, a testimonial cannot carry a claim the seller could not prove directly, and “results not typical” does not fix a dramatic one (FTC wording rules CX‑01, CX‑02).
05
The offer, decided in one look.
Yours: the offer page shows one plan only, a monthly subscription; the quiz page shows no price.
Ours: your three plans side by side, each with its price and saving: Start & save at $71.55 a month, the 3-month plan at $67.57 a month, and Try it once at $79.50. The delivery and billing terms of the chosen plan sit right under them, “Modify, skip, or cancel anytime” sits directly above the button, and one buyer’s line sits inside the longer plan.
Why: the price options belong together with the terms right under them, and “cancel any time” before the button (Shopify’s own guidance, C4‑07, C4‑08); a review line beside the higher-priced choice (C4‑02).
06
Built for the phone, and for the desktop too.
Yours: 20 to 24 phone screens per page; on a desktop the phone column sits in the middle of the screen.
Ours: on a phone, the story page’s first screen already holds the name, the rating, the price, the button and the guarantee; the offer page opens with the plans, and a bar with the price, the guarantee and the button stays in reach. On a desktop, the pages use the full width in your colours and type.
Why: most landing-page visits come from phones (83% in Unbounce’s data, all industries: C1‑07), and most viewing time is spent in the first two screens (57% above the fold, 74% in the first two, in an eyetracking study on desktop: C1‑06).

The figures above are from your two campaign pages as we saw them on 8 November 2026 (one visit each, nothing clicked), and from your product page as saved on 7 November 2026. The market figures are other companies’ published data and say how pages tend to behave, not what yours will do.

Your funnel, rebuilt.
Story, quiz and offer, linked.

Why this funnel, and how we would test it:

A story page in front of the offer is a bet, not a rule. Here is why we make it, which angle and which quiz we chose, and how we would prove it with your own traffic before calling it better.

01
Why a story page first.
A visitor from a cold Meta ad does not know Tributyrin-X yet. The story page warms her up before the product: the problem in her own words, how the capsule works, two buyers, then one next step. Readers who are ready skip it with one link.
Why: Demand Curve’s guidance for cold traffic is to “warm up cold traffic with storytelling and persuasive copy to build trust before redirecting to a product page”, with Ritual as the example (C1‑11). The same page warns that this adds a step where most advice removes steps, so we do not assume it wins: we test it.
02
The test: the funnel against your product page.
One ad, one angle, two groups of cold visitors. Group A goes to the story page and on through the quiz to the offer. Group B goes straight to your Tributyrin-X product page. Same ad, same prices, same guarantee: the only change is the path, so the result has one cause. We compare revenue per visitor and orders per group. Visitors who already know you, from email or as returning customers, stay out of this test and go to the offer or your product page.
Why: one change per variant, so a result has a cause (Shopify, C5‑03); warm traffic goes to the product page (C2‑02).
03
The size of the test, set before launch.
How many visitors each group needs is computed from your own conversion rate before the test starts: visitors per group = 16 × rate × (1 − rate) ÷ (the change we want to detect)². That gives an 80% chance of seeing a real change, at 95% confidence. PostHog’s own example: at a 10% conversion rate, seeing a 20% change takes 3,600 visitors per group. At 2%, a rate we use here only as an example, the same change takes 19,600 per group: the lower the rate, the more traffic a fair test needs.
Why: PostHog’s documented formula (C5‑01). A smaller change needs more visitors and more time, so the length of the test comes from your traffic, not from the calendar.
04
Orders, not days, decide when we read it.
Your Tributyrin-X product page already loads the Intelligems A/B testing app, so the test runs on a tool already on your store. Intelligems’ own guidance: about 300 orders in each group to see a 10% change in conversion with 90% confidence, and before that “a single conversion creates huge percentage swings”. So we run two groups, not more, unless your traffic gives every group 300 orders. The stop point is fixed before launch and the result is not read early: looking ten times at a running test means a result has to pass at 1% to keep a real 5% error rate. A winner is tested again later, because the season can move the result.
Why: Intelligems’ documentation (C5‑02, C5‑04); Evan Miller’s “How Not To Run an A/B Test” (C5‑05).
05
One angle now: “you’ve tried to fix it for years”.
The story page opens with your own headline and line from your product page: “Stubborn gut problems have finally met their match” and “…that you’ve tried to fix for years”. It speaks to the long-term sufferer, who has tried other things before. Your two campaign pages each speak to one situation: constipation after 45, and the gut in menopause. We build one angle first and build it fully; the next angles, yours among them, reuse the same page with their own headline and opening line.
Why: the ad and the page must say the same thing (C1‑01); build the strongest page for the biggest source first, then branch (C3‑05); keep the number of versions small at the start (C3‑07); one section order for every audience, only the header changes (C3‑01, C3‑03). Your own words also mean the page makes no claim your store does not already make.
06
The quiz: one product, so it chooses the plan.
Five questions about Tributyrin-X only: what bothers her most, for how long, what she has tried, how her gut takes new supplements, and whether she has taken it before. Your own plan notes decide the result: a first start gets Start & save, “Best for dialing in your dose”; a returning buyer gets the 3-month plan, “For once your dose is established”. The result names her concern, shows your dosing help and a buyer whose words match, and carries the button. No email is asked, no dose is suggested, and the quiz says plainly that it does not diagnose.
Why: a quiz earns its place when its result carries the recommendation and the buy button (C2‑04). With one product there is no product to choose (C2‑07), so the quiz makes the choice your product page asks the buyer to make: the plan. Whether the quiz itself helps is part of what the test in row 02 measures.

The formulas and thresholds are from PostHog’s, Intelligems’ and Evan Miller’s published guidance. The 2% rate is an example, not your figure. No result is promised: the test is there to find out.

Case studies.
Every case, in one list.