Making you hungry before you reach the bottom of the page

What surfaced

• 01

  Local bakery sites were hard to navigate and didn’t read as trustworthy. The bar was low.

• 02

  People didn’t want to get lost in a catalog. They wanted the shortest path to ordering.

• 03

  A new brand borrows no reputation. Trust had to come from the food and from proof that other people already buy there.

• 04

  Catering leads die from slow replies and long forms. That’s what people cited for giving up on competitors.

• 05

  Almost all traffic would be mobile, and nobody had checked whether a long page works on a phone.

Underneath both: nobody arrives at a new bakery’s site with patience or with prior trust. Both have to be won in the first few seconds, and not with brand content, but by removing steps and showing that other people already buy there.

Every insight produced a decision

What the interface is built on

• The photo doesn’t illustrate, it converts

  Images of desirable food produce measurable neural and physiological responses. Which is why every product shot is cut out: with no background competing, there is nothing else to look at.

  Spence et al., Brain and Cognition, 2016

• Colour sets the flavour expectation

  Colour and basic taste associate consistently: red and pink with sweet, brown with roasted. Beikit’s palette lives entirely in that register.

• Curves taste sweet

  Rounded forms read as sweet, angular ones as bitter. The system is obsessively curved: squircle radius, waves, a sun-shaped isotype, script type. The one angular element is the starburst, reserved for novelty and never used on product.

• Aesthetics substitute for reputation

  An attractive interface reads as more trustworthy, and that perception transfers to the product. The von Restorff effect is why the “Bestseller” badge works: it is scarce.

On that foundation

01

Hero CTA hierarchy

“Pide ahora” primary, “Ver carta” secondary, while catering doesn’t exist as a destination yet. The Google rating sits directly beneath: social proof at the decision point, not buried in a footer.

02

Cards that sell appetite, not information

On hover the photo scales up and the whole card takes the product’s colour: a cookie goes brown, a matcha cheesecake goes green. Not decoration. It sets a flavour expectation before the user reads a word.

03

Three cards per category, not nine

Building the catalog made it obvious it was too long. Hick’s Law and the choice-overload literature agree: nobody decides better with nine flavours competing before they have picked a category. The default caps at three and expands on demand.

Hick 1952; Iyengar & Lepper, Journal of Personality and Social Psychology, 2000

04

“Bestseller” badges

For a three-week-old bakery with no recognition, a “Bestseller” tag does the job an established brand gets free from word of mouth. It borrows the crowd’s validation before the crowd is visible on the site.

The check I ran with Juan and Anna before shipping: the badge goes only on demonstrably most-ordered items, never on what they want to push that week. That is the line between using social proof and faking it. Category order is their call; the badge and the cap are mine.

05

Symmetric cookie buttons: the pattern I chose not to use

European regulators have spent two years fining banners that gave “Accept” more visual weight than “Reject”. Beikit’s gives both buttons the same size: same padding, same type, and the only difference is a fill.

GDPR Article 7 requires it, and it is consistent: every other persuasive pattern here pushes someone toward what they came to do. A dark pattern in a consent banner would push them away from a decision about their own data.

1. 
   01

   Hero CTA hierarchy

   “Pide ahora” primary, “Ver carta” secondary, while catering doesn’t exist as a destination yet. The Google rating sits directly beneath: social proof at the decision point, not buried in a footer.

2. 
   02

   Cards that sell appetite, not information

   On hover the photo scales up and the whole card takes the product’s colour: a cookie goes brown, a matcha cheesecake goes green. Not decoration. It sets a flavour expectation before the user reads a word.

3. 
   03

   Three cards per category, not nine

   Building the catalog made it obvious it was too long. Hick’s Law and the choice-overload literature agree: nobody decides better with nine flavours competing before they have picked a category. The default caps at three and expands on demand.

   Hick 1952; Iyengar & Lepper, Journal of Personality and Social Psychology, 2000

4. 
   04

   “Bestseller” badges

   For a three-week-old bakery with no recognition, a “Bestseller” tag does the job an established brand gets free from word of mouth. It borrows the crowd’s validation before the crowd is visible on the site.

   The check I ran with Juan and Anna before shipping: the badge goes only on demonstrably most-ordered items, never on what they want to push that week. That is the line between using social proof and faking it. Category order is their call; the badge and the cap are mine.

5. 
   05

   Symmetric cookie buttons: the pattern I chose not to use

   European regulators have spent two years fining banners that gave “Accept” more visual weight than “Reject”. Beikit’s gives both buttons the same size: same padding, same type, and the only difference is a fill.

   GDPR Article 7 requires it, and it is consistent: every other persuasive pattern here pushes someone toward what they came to do. A dark pattern in a consent banner would push them away from a decision about their own data.