Take a ten-minute link dose: free lookup, $100 credit on sign-up.

RankDose
RankDoseGet $100 free

Checklist

Traffic drop backlink check: ten minutes, one clear answer

Organic traffic fell and everybody has a theory. Breathe first. This checklist uses link data to answer one narrow question: did the links change, or should you look somewhere else?

White card model road with a clear break indicating a disconnected link
Before the first call

Write down the date the drop began, to the day if you can. Every step below compares that date with something else, and a vague date makes every answer vague.

Ten minutes, in this order

  • Minute 1: open Search Console and look for a manual action or security message. If there is one, stop here and deal with it first.
  • Minute 2: compare your drop date with Google's public update history, for example the March 2024 core update or the August 2025 spam update.
  • Minute 3: check whether the lost clicks sit on a few pages or across the whole site. A few pages points at content or indexing.
  • Minute 4: run the free overview for your domain and note referring domains and DR.
  • Minute 5: open your notebook line from a month ago and subtract.
  • Minutes 6 to 8: pull referring domains for 2 credits and compare names with last month's file.
  • Minute 9: run one compare call, 1 credit, against your usual rivals to see whether the whole niche moved.
  • Minute 10: write one sentence naming the most likely cause, and one thing you will not change yet.

Reading the answer

A simple decision tree in table form. The numbers are examples for judgement, not Google rules.
What the link numbers showWhat it likely meansNext move
Referring domains flat or rising, DR steadyLinks are probably not the causeLook at content, indexing and the update date
Referring domains down by about a fifth or more since last monthSeveral linking sites vanished or movedRead the lost names, ask the real ones to restore the link
Your numbers flat, rivals clearly upYou stood still while others movedRead the competitor gap table
A sudden rise of odd domainsPossible spam burstFollow the order on the alarm-line page
Numbers unchanged since your last lookIndex probably has not refreshed yetWait for the monthly refresh before concluding anything

Three calls, written out

01

Free overview

Request action=overview for your domain with no key. Note the date, DR and referring domains, then compare them with your notebook line from last month.

02

Referring-domain pull

Request action=refdomains with per=100 and order=desc. It costs 2 credits whatever the number of rows, so there is no reason to skimp on the page size.

03

One compare call

Add your usual rivals to a single compare call at 1 credit. If every domain slipped together, the cause is probably bigger than your own links.

Wooden scale balance comparing a card block against a stone

How to read the date of the drop

The date is the spine of this traffic drop backlink check. Everything hangs on it. Every step compares it with something else, so a careless date makes every answer careless too.

The check is only as good as the notes you wrote before opening any tool, so write them first and open the tools second. Write first.

Which date counts when the drop was gradual?

When traffic slid over several weeks rather than falling off a cliff, pick the first week it dipped clearly below its usual range. Choose your anchor. Use that week as the anchor, and record the earlier weeks as a band rather than a single day.

Gradual declines are hard to date, so choose one line and stick with it. Be consistent. Consistency matters more than precision here.

If the slide continued for a month, the anchor week can move. Keep the first anchor in the notebook anyway, so the change is visible rather than hidden. Keep the first anchor.

What if the drop started on a weekend or a holiday?

Weekend numbers look soft in many accounts. Like with like. Compare like with like, Saturday with Saturday, and do not measure a Monday against a Sunday. A holiday dip can look like a crisis on a chart.

Ask whether the same period last year showed a similar dip before you call it a change in the site. Check last year first.

Keep the calendar in view as you read the numbers, and note any event that falls inside your window. Write the events down.

Does the update history prove the cause?

It proves only that a change happened near that date, which is weaker than it sounds. Dates alone prove little. Several things can change in the same week, and an update announcement is a clue to test rather than a conclusion. Keep it in your notes as one candidate among several.

It is a clue. Nothing more.

Four notes to write before the check

Four short notes give the rest of the check a fixed starting point.

01

The first bad day

Write the first day the numbers looked wrong, and the last normal day before it. Two dates give you a window, which is more honest than a single date.

02

Where the loss sits

Note which pages or sections lost the most clicks over the same date range you used before. Losses usually cluster, and a cluster points somewhere more useful than the whole site does.

03

What changed on your side

List the changes made in that window: pages published or edited, redirects, theme or plugin updates, and any links you asked for. Even small edits belong on the list.

04

What the rivals did

Check whether the rivals in your usual compare moved in the same window. Shared movement points away from your own links.

White card model bridge showing two distinct road levels

When a drop recovers and then falls again

A short recovery is confusing, and it can make you doubt the whole check. Confusion is normal here. It is still useful information, because it tells you something changed twice.

Why did traffic bounce back and then slide again?

A short recovery often means one cause was removed while another stayed active. Look for the pattern. Look at which pages moved in both directions, since the same pages usually tell the story twice. Then log the second drop as a new event with its own date.

Treat the two drops separately in the notebook, even if they feel like one problem. Keep them apart.

A second drop does not cancel the first. Log both. Each needs its own explanation and its own line.

Is it safe to change several things at once?

It makes the next drop much harder to read. Too many changes blur it. If you edit pages, change links and move the sitemap in the same week, the notebook cannot say which change helped or hurt. Change one thing, wait for the data, then change the next.

Slow edits look boring on paper. Boring is useful. They also leave you a trail you can follow.

Should I wait for the next core update before doing anything?

Not always. A core update can take a while to settle, and a drop that began just before one may reverse. Waiting is reasonable when the evidence is weak. It is less reasonable when a clear fault, such as a broken template or a blocked page, is sitting in front of you.

Fix the faults you can see, and watch the trend for the rest. Watch the rest. The notebook makes that split easier to keep.

Signs the cause is probably not links

  • Referring domains stayed steady across your last two readings.
  • The lost clicks sit mostly on content pages, not on one linked article.
  • Your usual rivals moved in the same window.
  • Search Console shows no manual action or security message.
  • The drop began before any new links appear in your notebook.

Choosing the next move after the check

The check ends with a decision, and a decision needs one clear next step. Pick one. The two situations below cover most of what the ten minutes reveal.

What if links look fine but traffic is still down?

Then the cause is probably somewhere other than your backlinks, and that is useful to know. Look at the pages. Move your attention to the pages that lost clicks, and check whether they still answer the search the way they once did. Compare them with the pages that held steady.

Write the next step in the notebook as one sentence. One step only. A single clear step is easier to act on than a list of six.

What if the links clearly moved?

If referring domains fell by a real amount, read the lost names first. Names are evidence. Ask whether those sites still exist, whether the pages moved, or whether someone simply removed a mention. Many losses are ordinary, and some can be fixed by asking the linking site to restore the link.

Do not send a flood of requests. Quality beats volume. Pick the few lost links that matter most, and keep a record of each one you ask about.

Why we do not chase links first

Most drops are not about links. Links are one suspect. Sites lose clicks when a core update reshuffles what searchers see, when pages fall out of the index, or when an AI summary answers the question above the results.

A link check is still worth ten minutes. Ten minutes is cheap. It is quick and cheap, and it lets you cross one suspect off the list, which feels much better than a fortnight of guessing at three things at once.

Be careful with recovery stories. Traffic often returns only after a later core update reassesses the site, and nobody outside Google can give you a date. Our data cannot speed that up, and any provider who promises it is selling hope.

Write the date first. Then wait for the next refresh before you judge a single week, because one flat reading proves little and four in a row prove a lot.