Guide5 min readUpdated 28 August 2026

Why did my website traffic drop?

Short answer

Almost always one source moved and the rest did not, so open the sources report and compare periods before you read the total at all. Rule out your own tracking first: a deploy that drops the script produces a fall that looks exactly like a traffic problem.

The tables below are one demonstration site over two weeks. The figures are illustrative; the reports and the comparison are the ones you already have.

Check the date range first

Before diagnosing anything, make sure you are comparing like with like. A month with four weekends against a month with five will show a drop that means nothing. So will a range that includes a public holiday in your largest market.

Set the comparison to the previous period of the same length, or to the same period last year if your traffic is seasonal.

Split by source before anything else

Sources are where the answer usually is. A drop concentrated in one referrer means a link fell off a busy page, a thread stopped being read, or a newsletter went out last month and not this one. Those are not problems with your site.

A drop across organic search is a different story and points at rankings, a de-indexed page, or a change in how many people are searching for what you cover.

  • One referrer falling and the rest flat: the link cooled off, expected.
  • Organic search falling while direct holds: check rankings and indexing.
  • Everything falling by a similar share: check tracking before anything else.
Traffic · sources/traffic
SourceThis weekLast weekChange
reddit.com / referral4021,180−65.9%
google / organic3,4103,502−2.6%
Direct / none1,9051,860+2.4%
newsletter / email640612+4.6%

What it told us. One referrer lost two thirds of its visits and everything else held. That is roughly 780 of the 800 visits the total lost, and none of it happened on the site.

OpenTraffic → SourcesRealtime

Rule out a tracking problem

If every source fell by roughly the same proportion on the same day, suspect your own instrumentation before you suspect the internet. A deploy that dropped the script tag, a consent banner change, or a new route that never loaded analytics will all look like a uniform collapse.

The realtime view is the fastest check. Load a page yourself and see whether it appears.

Then look at pages

Once you know which source moved, find the pages it used to land on. A single article losing most of its traffic is a ranking or referral story. A whole section losing traffic evenly is more likely a technical change, such as a redirect or a robots rule that went out with a release.

Traffic · pages/traffic
PageThis weekLast weekChange
/blog/tracking-without-cookies210990−78.8%
/1,4041,388+1.2%
/pricing612604+1.3%
/docs/quickstart430441−2.5%

What it told us. One article, and it is the one the referral pointed at. A whole section falling evenly would have been a redirect or a robots rule instead; a single page falling is a story about the link, not about the site.

OpenTraffic → Pages

Decide whether it matters

Not every drop needs fixing. Losing low-intent referral traffic while conversions hold flat is a good trade, and chasing the visitor count back up would waste the month. Check conversions and average visit time alongside the total before deciding it is a problem at all.

Where that leaves you

The week lost about 800 visits and every one of them came from a single referral to a single article. Nothing on the site broke, conversions held, and there is no fix to ship — which is a different outcome from the same total falling because organic search moved. Getting to that distinction took two reports and one comparison.

Where this evidence is weak

One week against one week, so a single quiet day moves every figure above. And VisitWatch holds no cross-site data, so a referrer that removed its link looks identical to one that lost its ranking — that distinction comes from a search console, not from here.

Questions this guide answers

4
  • Why did my website traffic drop?Split by source
  • How do I tell a tracking problem from a traffic problem?Rule out tracking
  • Which date range should I compare against?Check the range
  • Is every traffic drop worth fixing?Decide whether it matters
Free14 days, no card

Do this with your own numbers

Every table above is a standard report over a standard range. Nothing here was built for the example.

14-day free trial. No credit card required.