You can usually tell a templated trade website from across a crowded café. It has a smell. Four CSS decisions are responsible for 80% of that smell. Fixing them is the cheapest way to take a site from "looks like everyone else" to "looks built". This is the list.
1. Border radius (the giveaway)
Templates use either 0px (brutalist) or 8px (default Bootstrap) on everything. Both look like the designer didn't make a choice. We pick a single radius value per site - typically 10px (0.625rem) - and apply it consistently to every interactive element. Cards: 10px. Buttons: 10px. Input fields: 10px. Image containers: 10px.
The consistency is what reads as "designed". The specific value matters less than the discipline.
What we never do: a different radius on cards vs buttons vs inputs. That's the template tell.
2. Font pairing
Templates default to Inter for everything, or Open Sans for everything. Both are fine; neither has character. We use a system font for the body (faster, no FOIT, native to the OS) and pick a specific display font for the headings.
On most Built Local builds the heading font is Inter at 600-700 weight, tight letter-spacing, generous line-height. The body is the OS system stack. The pair feels intentional without being noisy.
What we never do: three or more fonts. Two is plenty. Three is showing off.
3. Photo treatment
The single biggest "templated" tell: photos that all look different from each other because they came from different stock libraries. We process every client photo through the same pipeline:
- Slight saturation boost (+8%).
- Slight contrast lift (+5%).
- Consistent crop ratios (16:9 or 4:5 only - no square stock-photo defaults).
- AVIF compression to our 80KB above-fold budget.
- One unified colour grade across the site, even if photos were taken in different conditions.
The unified grade is the bit that does most work. Even mismatched photos from different days start to feel like they belong to one business.
4. Motion budget
Templates either use no animation at all, or use the pre-set "fade in on scroll" on every section. Both feel cheap. We have a motion budget per site: at most three discrete animated moments, each under 400ms, all triggered by user action rather than scroll.
The three we typically use:
- A 200ms transform on button hover/tap (lift + slight scale).
- A 300ms expand on FAQ accordion open.
- A 250ms fade-and-slide on mobile menu open.
That's it. No scroll-triggered fades. No parallax. No marquee. The site feels responsive to the user, not to a scroll position.
The composite effect
Each of these decisions on its own is small. The composite is the difference between "this looks like a site" and "this looks like *my* site". Customers don't articulate this - they say things like "the new site looks really professional" or "it doesn't look like a template". The CSS work is invisible; the perception is real.
The measurable bit
We can't easily measure "feels designed". We can measure the motion-budget decision. Removing scroll-triggered animations from a previously-animated site lifted average session duration by 11% in our A/B and dropped the bounce rate by 6%. People stayed because the site stopped distracting them.
What this means for UK local web design
If you're hiring a designer, look at three of their previous sites side by side. If you can spot four shared design decisions across all three (radius, type pairing, photo treatment, motion budget), they have a system - even if you don't love the specific decisions. If you can't, you're going to get whatever the template did by default.
Every Built Local site uses the same four decisions, tuned subtly per trade. Browse the demos page to see the pattern across different trades, or read the pillar guide for the broader design philosophy behind local web design for tradespeople.