Your website traffic suddenly dropped: what to check before you panic

You open your analytics on Monday. Last week was a spike. This week is close to zero. Almost all of the drop sits in one bucket: Direct.

This is a calm checklist for that moment. We have seen a pattern like this in our own numbers: a burst of Direct traffic, then a sharp fall. The data did not tell us the cause, so we did what we suggest here. We checked the plumbing first.

We are not going to give you a diagnosis. We cannot see your setup. We can give you an order of operations.

First, decide if the drop is real

A big fall after a big spike is often a return to normal, not a crash. The spike is the odd week. The quiet week may be closer to your real baseline.

Do this before anything else:

• Compare the quiet week to a longer window, such as the last 28 days, not only to the spike week.
• Look at the same weekday pattern from a month ago.
• Check whether the drop is spread across channels or sits in one.

If one channel explains almost all of the change, that is a clue. Direct is a catch-all bucket. Traffic lands there when the source is unknown, not only when someone types your address.

Second, check that tracking still works

A drop that looks like a cliff can simply be a tag that stopped firing.

Look at the tag itself

Open your site in a private window. Use your browser's developer tools or your analytics tool's real-time view. Load a page and see whether your own visit shows up.

If it does not, check:

• Did anyone change the site template, theme or hosting recently?
• Was the tracking snippet removed or moved by a deploy?
• Did a tag manager container get unpublished?

Look at consent

If you added or changed a consent banner, that can cut recorded sessions overnight. Visitors who decline are not counted, and a misconfigured banner can block everyone by default.

Ask one question: did anything about consent go live around the date the numbers fell? If yes, that is your first suspect.

Third, look for a one-off burst

Now look at the spike week itself. Ask what it was made of.

• Did one page or one day account for most of it?
• Were the sessions very short, with one page each?
• Did sign-ups or leads jump at the same time, and then stop?

A sudden jump and a sudden stop, matched across traffic and sign-ups, can point to automated traffic or a bulk import. It can also be a real launch or a mention that ran its course. The numbers alone do not say which.

One practical check: look at what those leads did next. If a large batch signed up and almost none did anything afterwards, treat the batch with caution. Do not count it as demand until you know where it came from.

Fourth, check your other sources

Your site analytics is one view. Cross-check it against something that does not depend on the same tag.

• Search Console shows impressions and clicks from search, with no tag on your site.
• Your server or hosting logs show raw requests.
• Your email or CRM shows sign-ups, which are independent of page views.

If Search Console is flat while site analytics collapsed, suspect tracking. If both fell, look at what changed on the site or in your rankings. Keep in mind that small numbers swing a lot. A move from 28 clicks to 13 is real arithmetic but a weak signal.

Fifth, write down what you changed

This sounds dull, and it is the step founders skip. Make a short list of everything that changed in the two weeks before the drop:

• Site or theme updates
• New or changed scripts
• Consent or cookie settings
• DNS, domain or redirect changes
• Campaigns that started or ended

Most mysteries shrink once the list is on a page.

What to do while you do not know

You may not find the cause in an afternoon. That is fine. Here is how to act in the meantime.

• Do not report the number to anyone as a trend. Say the data is under review.
• Do not make decisions about spend or strategy from one odd week.
• Do not rely on a single tool. Pair it with a second source.

If you are showing traction to anyone, a team, a partner, an investor, stick to what you can verify. A number you cannot explain is worse than a smaller number you can.

Add a signal that does not depend on your site tag

There is a second kind of data worth having: what happens after you send someone your link.

Site analytics tells you about anonymous visitors. It does not tell you whether the one person you emailed on Tuesday actually opened what you sent. For that you want a link made for that one person.

On StartupCorners, you get one clean link for your startup, built from your website. You can also create individual tracked links for different recipients. Each one shows opens, time spent and tabs viewed. That is all it shows. We do not track location, and we do not promise replies from anyone.

It is a narrow signal, but it is a steady one. If your site traffic looks strange, your sent links still tell you whether the people you wrote to looked at your page.

To be clear about where we are: StartupCorners is free and in beta. The page is built from your company website, which is why we ask for a work email. It is not a replacement for your analytics. It is one more place to check when you want to know if a specific message landed.

A short version to keep

When traffic falls off a cliff:

1. Compare against a longer window, not only the spike.
2. Test your tag in a private window.
3. Check whether consent changed.
4. Look at what the spike was made of.
5. Cross-check with Search Console and your CRM.
6. List every change from the last two weeks.
7. Hold off on conclusions until two sources agree.

Most of the time the answer is dull. A tag moved, a banner went live, or one burst ended. Dull is good. It means you can fix it.

Create your link

If you want one current link you control, and a way to see when it is opened, you can create yours for free. It takes your website and builds the page from it. Create your link.