Why did my organic traffic drop? A step-by-step diagnosis

Your organic traffic is down. The dashboard is red, someone important has noticed, and the question lands on your desk: what happened?

I have been on the receiving end of that question for twenty years, at Mister Spex, at Patient21, and everywhere in between. The instinct is to start guessing, "maybe it's a Google update?", "maybe the new release broke something?". Guessing is slow and it is usually wrong. There is a better order to work through, and it starts with the least glamorous suspect.

Step 1: Make sure the drop is real

Before you debug the site, debug the measurement. A surprising share of "traffic drops" never happened. The traffic was fine, the tracking broke.

Check, in this order:

• Did the GA4 tag stop firing? A new consent banner, a Google Tag Manager publish, or a deployment that removed the snippet will all flatline your numbers while real users keep visiting.
• Is the drop suspiciously clean? Real traffic declines are messy. A measurement break is often a perfect cliff: full volume one day, near zero the next, across every page at once.
• Does one channel hold steady? If direct and paid look normal but organic cratered, that points to a real organic problem. If everything dropped together on the same day, look at tracking first.

If the numbers are genuinely down, move on. If they are not, you just saved yourself a week.

Step 2: Pin the exact date

Find the day the line bent. Not the week, the day. Almost every real cause leaves a date stamp, and the date is what lets you match the drop to a cause.

Then classify the shape:

• A cliff (sudden, overnight) usually means something changed: a deployment, a tracking break, a manual action, a botched migration.
• A slope (gradual decline over weeks) usually means competition, decaying content, slow indexing loss, or seasonality.
• A step down that holds often lines up with a Google algorithm update.

The shape tells you which list to work through next.

Step 3: Work the usual suspects, in order

Technical and deployment changes

The most common cause of an overnight cliff is something your own team shipped.

• A noindex tag or a robots.txt rule that went out with a release
• A site migration, URL structure change, or redirects that broke
• A template change that stripped internal links, titles, or structured data
• A performance regression that tanked Core Web Vitals

Ask the obvious question: what did we deploy around that date? This is where most drops are solved, and also where most teams waste days, because nobody can remember what shipped when.

Google algorithm updates

If the drop is a step that holds and you ruled out your own changes, check whether Google ran a core or spam update around that date. Google publishes confirmed updates on its Search Status Dashboard. Match the date. If your drop starts on the day a core update finished rolling out, you have your answer, and a different project ahead of you.

Lost rankings and SERP changes

Even with no algorithm update and no deploy, you can lose traffic to:

• A competitor who overtook you on a high-volume keyword
• An AI overview or featured snippet that now answers the query without a click
• Lost backlinks that were propping up a key page

Pull your top landing pages and check rank and impressions for each, not just sitewide totals. A drop is almost always concentrated in a handful of URLs.

Seasonality and the outside world

Some drops are not your fault and not a problem. Compare year over year, not just month over month. A travel site in January and a tax tool in May move for reasons no deploy log will explain. Weather, holidays, and news cycles all count.

Step 4: Cross-reference, do not guess

Here is the pattern behind every fast diagnosis I have ever done: put the drop date next to a list of everything that changed around it. The deploys, the GTM publishes, the campaign launches, the algorithm updates, side by side on one date. The cause is almost always sitting right there. The slow part was never the analysis. It was reconstructing what happened, weeks later, from memory and Slack scrollback.

This is exactly the gap CoNote closes. It keeps an automatic, shared timeline of what changed across your stack, deploys, tag manager versions, campaigns, config changes, so when traffic moves you open one page, find the date, and read the reason. Know why the numbers moved, without the archaeology.