What Bounce Rate Actually Counts

A bounce is simply a session where someone arrived on a page and left without triggering anything your analytics counts as an interaction. In most modern dashboards that means no second page, no form start, no phone tap, no video play, nothing. They landed, they looked, they went.

That is all the number tells you. It does not tell you whether they were satisfied, bored, confused, or simply done. A visitor who found your opening hours in eight seconds and left happy is counted exactly the same as someone who landed on a broken mobile layout and hit the back button in frustration.

Before you read too much into it, remember that "normal" varies wildly by sector and by page type. A local service homepage might sit comfortably at 45–60%. A single blog post answering a quick question might sit at 80% and be working beautifully. Benchmarks from other industries are close to useless for your business.

Start With Page Intent, Not the Percentage

Ask a simple question about the specific page: what did the visitor come here to do?

• One-answer pages. Opening hours, delivery costs, parking info, a recipe, a how-to guide. If people get the answer and leave, that is success. A high bounce rate here is a sign the page is doing its job.
• Decision pages. Service pages, pricing, product pages, contact pages. These should push people onwards. A high bounce rate here deserves a closer look.
• Entry pages. Most visitors land on a page that is not your homepage. If that page does not clearly offer a next step, bouncing is the rational choice.

Signs of a healthy bounce include a decent stretch of time on the page, some scrolling, and the occasional tap on your phone number or email address. Signs of trouble include five-second exits, no scrolling at all, and a high exit rate on a page that sits directly on your conversion path.

Traffic Source Changes the Whole Story

A bounce rate is meaningless without knowing where those visitors came from. Segment by channel before you judge anything.

• Organic search to a blog post or guide tends to bounce higher, because people are often researching rather than buying. That is expected.
• Paid traffic to a purpose-built landing page should bounce lower, because you control the promise and the page. If it does not, your ad and your page are telling different stories.
• Email usually produces the lowest bounce rates, since your list already knows you.
• Social can be noisy. People tap, skim, and move on. A high figure here often says more about the platform than about your website.

A 90% bounce rate from a light-hearted social post is a completely different animal from 90% on your main service page from people searching for exactly what you sell.

Time on Page and Engagement Tell You More

Look at average engagement time alongside the bounce figure. Under ten seconds usually points to a mismatch between what the visitor expected and what they found. Two or three minutes with a bounce at the end often means they read everything and left satisfied.

Check scroll depth and any events you track. Did they reach the pricing table? Did they start the enquiry form? Did they watch the video? A page with a high bounce rate but strong scroll and event data is usually fine.

One caution: small numbers lie. Do not overhaul a page based on thirty sessions. Give it a few hundred before you draw conclusions, and remember seasonal swings. A quiet week in August is not a trend.

A Sensible Check Before You Change Anything

When a figure does look off, work through this in order rather than panicking:

• Confirm the tracking works. A broken event or a missing tag makes a perfectly good page look like a disaster.
• Compare the page to your own site average, not to an industry benchmark.
• Check the intent of the page. Is it meant to be a quick answer or a stepping stone?
• Check the traffic source. Segment by channel and by device.
• Check time on page and events. Short visits with no interaction are the real warning sign.
• Read the page as a stranger. Does the headline match what brought them there?

Only when a page has low engagement time, relevant traffic, and a genuine job to do should you start editing.

When It Is Worth Acting

The clearest case for action is a page on your conversion path with high bounce, low engagement time, and steady traffic from people who clearly want what you offer. That combination usually means one of four things: the headline does not match the search intent, the key information sits too far down the page, the next step is not obvious, or the page loads slowly on a phone.

The fixes are rarely dramatic. Tighten the headline so it echoes the search phrase. Move prices, coverage area, or availability above the fold. Make the call-to-action impossible to miss. Cut heavyweight images that stall on mobile connections. Add a couple of internal links to genuinely useful pages.

Change one thing at a time and leave it a fortnight before judging the result. Bounce rate is a diagnostic tool, not a score out of ten. Used carefully, it points you towards pages worth improving. Used anxiously, it sends you rewriting pages that were never broken in the first place.