Semrush Organic Research, and why you should stop starting from a seed keyword

Most keyword research starts with a seed word and a guess. You type something in, get a database dump, and spend an afternoon deciding which of 780,000 rows might be worth writing about.

Organic Research inverts that, and it is the reason I open it before the Keyword Magic Tool almost every time. You type in a competitor. You get the keywords they already rank for. Somebody has done the guessing, published the pages, and survived contact with the SERP.

Why competitor-first beats seed-first

The argument is on camera and it’s the right one. Seed-based research reinvents the wheel: you invent seed terms, trawl the database, and filter down to candidates whose demand you’re still inferring. Competitor-based research skips straight to keywords with three things already proven.

Demand exists. Somebody ranks and gets traffic for it.

The format is visible. You can open the URL that ranks and see whether Google wants a comparison table, a tutorial, or a definition. Seed research tells you a keyword exists, never what shape the answer takes.

The difficulty is real, not modelled. If a site with authority comparable to yours ranks for it, you know it’s reachable. That is worth more than any difficulty score.

The filter, which is the whole job

The raw output is as unusable as every other Semrush report. In the recording it opens at 2.9 million keywords for one domain and gets filtered down to 2,300. That is the work.

Order that gets there fastest:

1. Position band. Positions 1 to 20 only. Anything below that is a keyword your competitor is also failing at, which tells you nothing about whether it’s winnable.
2. Volume floor. Whatever “worth a page” means on your site.
3. Exclude their brand. Their company name and product names are unwinnable and they clog the top of every volume-sorted list.
4. Intent. Commercial where you sell, informational for the cluster around it.

Four filters, and you are looking at a few hundred rows you can actually read.

The Declined tab is the cheapest opportunity in Semrush

Almost nobody uses this, and it is the single best tactic on the page.

Organic Research splits movement into new, improved, declined and lost. Everyone looks at improved, because watching a competitor win feels like research. Look at declined instead.

A keyword a competitor is losing means the page that used to satisfy that query is getting weaker, whether through neglect, staleness, or Google changing what it wants. You are not attacking a page at its peak. You are stepping into a gap that is already opening, which is a completely different amount of work for the same reward.

Sort declined by volume, open the URLs that are slipping, and ask what the page stopped doing. That is half an hour that regularly produces a better content plan than a full seed-research session.

Where this fits, and where it stops

Organic Research handles one competitor at a time and produces observations. It is the middle step in a three-report sequence.

Before it, run Domain Overview to decide whether this competitor is even in your league. A thirty-second screen saves an hour of reading a list belonging to a site you cannot compete with.

After it, run Keyword Gap. Organic Research shows you what one rival has. Keyword Gap compares several against you at once and returns what you are missing, which is a to-do list rather than a finding.

And when there is no comparable competitor, because you are genuinely first in a category, this report has nothing to give you. That’s when you go back to seeds and accept the guessing.