Breadcrumb navigation is a short trail of links, such as Home / Coffee / Espresso, that shows where a page sits in a site's hierarchy and lets visitors jump back up a level in one click.
It is a secondary aid that helps people find their way.
It is not a conversion lever: the one controlled breadcrumb test in our database moved conversion by +1.8% and never reached significance.
Sticky headers are a different story. They can help or hurt, and our tests split both ways.
This page covers what each pattern does, what usability research recommends, and what real A/B tests show.
Vasta
Blend Commerce
Clean CommitClean CommitWhat breadcrumb navigation is (and is not)
Nielsen Norman Group's breadcrumb guidelines by Page Laubheimer define breadcrumbs as a trail that shows the current page and its "ancestors."
The trail reflects the site's structure, not the pages a visitor happened to click through.
NN/g warns that a session-history trail "will quickly become long and confusing."
The same guidelines say breadcrumbs "do not replace the global navigation bar."
They sit alongside the main menu. NN/g also recommends a few basics:
- Start the trail with a link to the homepage.
- Show the current page as the last item, not linked.
- Skip breadcrumbs on flat sites with only one or two levels.
- Only include levels that are real pages a visitor can land on.
- On mobile, keep the trail on one line and the tap targets at least 1 cm by 1 cm, even if that means showing only the parent level.
Breadcrumbs also have a search side. Google reads BreadcrumbList structured data, with a position, name and item URL for each step, according to its breadcrumb markup docs.
Since January 23, 2025, Google shows only the domain on mobile results and keeps breadcrumbs on desktop results.
So the search payoff now comes from desktop results only.
Sticky headers: what the tests show
A sticky header (also called a fixed or persistent header) stays pinned to the top of the screen as the visitor scrolls.
It keeps the logo, menu, search and cart one tap away.
The cost is screen space, which matters most on a phone.
Our database has tests pointing both ways. Here are the ones where the sticky header itself was the main change.
| Test | Change | Result | Ran by |
|---|---|---|---|
| Sticky desktop navbar | Added a fixed desktop header with logo, menu, cart and search | +15.3% conversion rate, +14.0% orders, +9.6% revenue per visitor; not significant | Clean Commit |
| Removing the sticky nav | Took the sticky header away on Canada Welding Supply | +1.59% conversion rate, +3.45% revenue per visitor, +1.83% AOV | Vasta |
| Fishwife homepage | Sticky header with gift and fish category links, plus a new collections layout | +8.9% revenue per visitor, credited to the sticky nav (bundled test) | Fishwife in-house team |
| Westfund quote button | Added a "Get a Quote" button to the floating top nav on desktop | +10.82% quotes started, +28.84% new member joins | Insighten |
When adding a sticky header helped
Clean Commit's sticky desktop navbar test looks big at first glance: +15.3% conversion rate across 19,934 desktop visitors in 11 days.
But the primary metric, click-through, reached only a 69.5% probability of beating control.
The agency labeled it a positive signal, not a winner, and so do we.
Fishwife's homepage test credits +8.9% revenue per visitor to a sticky header with clear category links.
It ran from September to November 2025 alongside a second change, so treat the split between the two changes as the source's own attribution.
The clearest win came from what was in the header, not the stickiness.
Insighten's test for health insurer Westfund put a "Get a Quote" button in the floating nav, where it had only been in the homepage hero before.
Quotes started rose 10.82% and new member joins rose 28.84% on desktop.
When removing a sticky header helped
Vasta ran the opposite test for Canada Welding Supply: they removed the sticky nav and changed nothing else.
In a 50/50 split over 14 days that reached significance, conversion rate rose 1.59% and revenue per visitor rose 3.45%.
Vasta's write-up says the bar was "crowding product imagery and stealing taps meant for product cards on mobile."
That matches the risks NN/g lists in its sticky header guidelines: headers that are too tall, see-through backgrounds that make text hard to read, and headers kept sticky when visitors rarely need them.
How to design a sticky header worth testing
- Keep it short. NN/g advises starting with readable text and then cutting the header's height as far as you can.
- Use a solid background. NN/g says the header "must be an opaque color," different from the content behind it.
- Consider showing it on scroll-up only. A partially persistent header that slides back in when the visitor scrolls up saves space. NN/g suggests a slide-in of roughly 300 to 400 ms.
- Put a job in it. Search, cart and a primary call to action earn their space. A logo alone does not.
- Watch for clashes. On product pages, a sticky header can compete with a sticky add-to-cart bar or an announcement bar for the same screen space.
Pinned navigation doesn't have to be the full header.
Blend Commerce tested sticky subcategory pills above the fold on Stone Creek Coffee's homepage and reported +8% conversion rate and +19.41% collection view rate for that variant.
Breadcrumbs in ecommerce
On a store, breadcrumbs matter most for shoppers who land deep in the catalog from search or an ad.
A product page reached from Google has no "back" path, so the trail is how the shopper finds the rest of the category.
Hierarchy breadcrumbs vs. "Back to results"
Baymard Institute's breadcrumb research (Jamie Holst, December 2013) argues stores need two kinds.
Hierarchy breadcrumbs send shoppers to the parent category but drop their filters and sort order.
A history-based "Back to results" link returns them to the exact list they came from, with filters kept.
In Baymard's benchmark of 50 major ecommerce sites, 45% offered only one type and 23% had no breadcrumbs at all.
Baymard's suggested fix is a normal hierarchy trail with a "Back to results" link added at one end.
The research is old, so check how your own platform handles filtered lists before you assume it applies.
What the breadcrumb A/B test showed
Clean Commit added clickable Home / parent / current collection breadcrumbs to collection pages on an anonymized Shopify store.
After 45 days, conversion rate was up 1.8% and revenue per visitor up 2.2%.
The probability of beating control was only 70.9%, so the result is inconclusive.
That is a useful finding in itself. Breadcrumbs rarely hurt, so you can treat them as a navigation standard rather than something to test.
A store with low traffic would need a very long test to detect a lift this small (see testing with low traffic).
A note on terms: some lead-gen agencies call multi-step forms with a progress trail "breadcrumb forms."
Those test results are about form design, not site navigation, so we leave them out here.
Which navigation changes are worth testing
| Pattern | Typical effect in our tests | Test it or just ship it? |
|---|---|---|
| Hierarchy breadcrumbs | Small, inconclusive lift | Ship it on catalogs with three or more levels |
| "Back to results" link | No controlled test in our database yet | Ship it if filters reset when shoppers go back |
| Full sticky header on mobile | Removing one won at Canada Welding Supply | Test it, and look at mobile results on their own |
| Sticky header on desktop | Positive but not significant | Test it if search and cart are used often |
| CTA inside a sticky header | Clear win at Westfund | Test it on lead-gen and high-consideration sites |
Look at mobile and desktop separately. Clean Commit's compact navigation redesign finished at -0.3% conversion overall, but that hid +23.9% on desktop and -4.7% on mobile.
A sticky header that helps on a wide screen can cost space on a phone (more in our mobile CRO guide).
Pick a primary metric before launch. Navigation tests move clicks easily, and clicks can rise while orders stay flat.
Use orders or revenue per visitor as the deciding metric, and only call a winner once it reaches statistical significance.
For more on menus, mega menus and category links, see navigation menu tests.
For the wider picture, see our guides to ecommerce UX and conversion rate optimization.