What Is a Good Bounce Rate for a SaaS Marketing Site?

There's no single authoritative bounce-rate benchmark for SaaS marketing sites. As a rough, unattributed guide, many land somewhere around 40–60% — but that's a general estimate, not a published standard, and the figure swings hard with traffic source and with how each analytics tool defines a "bounce." The number worth acting on is your own, measured consistently.

Why a single benchmark doesn't exist

The metric isn't comparable across tools because the definition itself changed:

• Universal Analytics counted a bounce as a single-page session with no second interaction.
• GA4 dropped that and reports the inverse of an engaged session — a session that lasts 10+ seconds, fires a conversion, or has 2+ pageviews. A GA4 "bounce rate" and a UA "bounce rate" are different metrics with the same name.
• PostHog doesn't ship one fixed definition; you compute it from raw events, so you decide what counts as engaged.

Quote a bounce rate without stating the definition and it means nothing. That's why any "industry average" you read — including the 40–60% guide above — is a starting point, not a target.

Traffic source moves it more than your industry does

A SaaS marketing site rarely has one bounce rate; it has several, hidden inside an average:

• Branded and organic search tends to bounce lower — those visitors arrived with intent.
• Cold paid social tends to bounce high — curiosity clicks that leave fast.
• Referral from a relevant article sits somewhere in between.

If your blended bounce rate looks alarming, segment by source before changing anything. A 65% blended rate that's really 35% on search and 80% on paid social is two different stories, and only one is a problem.

Measure your own bounce rate in PostHog

Here's a defensible definition — a session with only one pageview — expressed in HogQL over the last 30 days. It groups events into sessions, then counts the single-pageview ones:

SELECT
  count() AS sessions,
  countIf(pageviews = 1) AS single_page_sessions,
  round(countIf(pageviews = 1) / count() * 100, 1) AS bounce_rate_pct
FROM (
  SELECT
    properties.$session_id AS session,
    countIf(event = '$pageview') AS pageviews
  FROM events
  WHERE timestamp > now() - INTERVAL 30 DAY
    AND properties.$session_id != ''
  GROUP BY session
)

Illustrative sample output:

sessions	single_page_sessions	bounce_rate_pct
7,300	3,580	49.0

A 49% result sits inside the rough guide — neither a red flag nor a reason for complacency. If you'd rather not treat a quick, satisfied visit as a bounce, tighten the inner query to also require no scroll past, say, 50% using properties.$prev_pageview_max_scroll_percentage. The point is that you own the definition, so state it whenever you report the number.

What to do with the number

Don't chase a lower bounce rate for its own sake. Establish your own baseline, segment it by traffic source and by landing page, and act only where a high bounce contradicts the page's purpose — a homepage that's meant to pull people deeper, a signup page that's meant to convert. Compare each page to its own history week over week, not to a borrowed benchmark.

If you'd like that analysis done for you — your funnel read from PostHog and the highest-impact fix shipped as a Pull Request — that's what Velyr does.