Loading tutorials…
Loading tutorials…
Keyword Magic Tool returns 30,000 keywords from a single seed. Most teams export the first 200, dump them into a spreadsheet, and never ship a piece of content from the list. This is the workflow specialists actually use — from seed to ranked content brief in under 5 hours.
Who this is forFounders, content marketers, and in-house SEOs running monthly content programs who pay for SEMrush Pro/Guru and use Keyword Magic Tool ad-hoc instead of as a workflow. If your content calendar is built on gut-feel keywords instead of a clustered keyword map, this is the reset.
What you'll need
Step 1
SEMrush → Domain Overview → enter your domain. Note your Authority Score. KD scores need discounting if your AS is below 40.
Open SEMrush → Domain Overview → enter your domain. The Authority Score (AS) appears in the top-left. AS 0-30 is new/small; 30-50 is established; 50-70 is competitive; 70+ is dominant.
SEMrush's Keyword Difficulty (KD) scores are calibrated for sites with AS 50+. If your AS is below 40, treat KD as 'absolute difficulty' minus a discount: KD 30 actually behaves like KD 45 for you. KD 50 behaves like KD 70.
Practical rule: at AS < 30, target only KD < 20 keywords for the first 6 months. At AS 30-50, KD < 35. At AS 50+, KD < 50.
Bookmark your AS — re-check it monthly. As your backlink profile grows, your usable KD ceiling rises and the keyword pool expands.
Step 2
SEMrush → Keyword Magic Tool → enter seed → set country → choose match type. Use 3-5 seeds, not 1 — broader topical coverage in less time.
Open Keyword Magic Tool (left rail under SEO). Enter your first seed keyword. Set country to your target market.
Switch through the four match types in the tabs: Broad Match (default, biggest list), Phrase Match (contains the seed phrase), Exact Match (exact word order), and Related (semantic siblings). Phrase Match is usually the highest-signal default.
Don't run one seed and call it done. Run 3-5 seeds covering your category — e.g., for a project management SaaS: 'project management software,' 'task management,' 'team collaboration tool,' 'workflow software,' 'gantt chart.' Each seed surfaces different long-tail clusters.
Use the left-rail filter panel: Volume (min 50), KD (your calibrated ceiling), Intent (Informational, Commercial, Transactional, Navigational), Word count (4-8 for long-tail), SERP features (check 'Featured snippet' if you want snippet-target keywords).
Save filtered keyword sets to a Keyword List (top-right Save button). Name lists by cluster — 'pm-software-bofu,' 'pm-software-tofu' — not 'List 1.'
Step 3
Intent column → filter to Commercial + Transactional for money pages, Informational for top-of-funnel. Strip questions, brand-match noise, and parasite SERPs.
SEMrush assigns each keyword an Intent label: Informational (learning), Commercial (researching options), Transactional (ready to buy), Navigational (brand or site search).
For top-of-funnel content (blog posts, guides): filter to Informational + Commercial.
For money pages (landing pages, comparison pages, pricing pages): filter to Commercial + Transactional.
Strip Navigational keywords (containing competitor brand names) unless you have a specific brand-comparison play.
Strip 'parasite SERP' keywords — queries where SERPs are dominated by Wikipedia, YouTube, Reddit, Quora, or major news sites. The SERP Features column flags these.
Strip questions you can't credibly answer (e.g., regulated industries with legal-liability questions). Use the Questions filter under Advanced filters.
A clean filtered list of 200-500 keywords beats a raw export of 5,000. Stop chasing volume; chase actionability.
Step 4
Keyword Magic Tool → select keywords → Add to Keyword Strategy Builder. The tool clusters by SERP overlap into pillar + supporting pages.
From Keyword Magic Tool, select 100-500 filtered keywords and click 'Add to Keyword Strategy Builder' (top-right).
Keyword Strategy Builder analyzes SERP overlap between keywords and clusters them automatically. Keywords that share 4+ ranking URLs are grouped into the same cluster.
Review the auto-clusters. Each cluster becomes one piece of content: the cluster's highest-volume keyword is the primary keyword, the others are supporting subheadings.
Tag clusters by funnel stage (TOFU, MOFU, BOFU) and content type (guide, listicle, comparison, how-to, definition).
Export the cluster map as CSV. This is your content calendar — each cluster = one article, prioritized by aggregate cluster volume × inverse KD.
Step 5
SEMrush → Topic Research → enter pillar keyword. Returns subtopics, questions, and headlines that map to real reader curiosity.
Open Topic Research (left rail under Content Marketing). Enter the pillar keyword from one of your clusters.
Topic Research returns: subtopic cards (related themes with volume), Headlines (high-engagement headlines from the web), Questions (real questions from People Also Ask + forums), and Related Searches.
For each cluster you're writing, pull the top 5 subtopic cards as outline sections.
Pull the top 10-15 questions as H2/H3 candidates. These map directly to People Also Ask featured snippets.
Cross-reference Topic Research questions against the Keyword Magic Tool 'Questions' filter — any overlap is a high-confidence H2.
Save the Topic Research export. This is your draft article outline.
Step 6
Combine cluster map + Topic Research output + competitor outlines into one brief. SEO Writing Assistant validates the brief in real time.
Open Google Docs or your CMS. Create the content brief: primary keyword, supporting keywords (from cluster), target word count (check top-ranking competitor lengths in Keyword Magic Tool → SERP Analysis), and the outline from Topic Research.
Use the SERP Analysis feature in Keyword Magic Tool: click the primary keyword → SERP Analysis tab → see top 10 ranking URLs, their word counts, referring domains, and authority scores. This tells you what the SERP rewards.
Set word count target to the median of positions 3-7 (not the top result — that's often an outlier).
Open SEO Writing Assistant (SEMrush left rail under Content Marketing). Paste your draft outline. SEO Writing Assistant scores the brief against the SERP and suggests semantic keywords and questions to include.
Assign the brief to a writer with a deadline. Track the cluster through publish + index in Position Tracking.
Step 7
Once content publishes, add the cluster keywords to Position Tracking with a cluster tag. Track ranking movement weekly.
Once the article publishes, open Position Tracking → Add Keywords. Paste the cluster's primary + supporting keywords.
Tag them with the cluster name (e.g., 'cluster-pm-software-comparison'). This lets you filter the dashboard to see how the whole cluster moves together.
First ranking signals appear in 2-6 weeks (faster on existing domains, slower on new ones). Don't panic if rankings don't appear in week 1.
If a cluster sits outside the top 50 after 8 weeks, the content may not match search intent. Re-check the SERP and consider a rewrite, not a retire.
If a cluster lands 5-15, it's promising — build internal links from related content and consider a backlink campaign to lift it into top 5.
Common mistakes
Ignoring Domain Authority when reading KD
What goes wrong: You target 'KD 30' keywords with a brand-new site (AS 15). You write 20 articles. After 6 months, average ranking is position 45. You blame SEMrush. The wasted content cost: ~$200-400/article × 20 = $4,000-8,000 of misallocated content spend.
How to avoid: Calibrate KD against your real AS. At AS < 30, target KD < 20. At AS 30-50, target KD < 35. As AS rises, the ceiling rises with it.
Exporting raw keyword lists without clustering
What goes wrong: You export 3,000 keywords from Keyword Magic Tool, drop them into a spreadsheet, and assign 'one article per row.' 30 articles in, you realize 20 of them target overlapping clusters and now compete with each other in the SERP. Keyword cannibalization tanks half your articles.
How to avoid: Always use Keyword Strategy Builder to cluster before assigning to writers. One cluster = one article. Track which keywords belong to which cluster.
Skipping intent filtering
What goes wrong: You write a transactional landing page targeting an informational keyword. The SERP rewards a 3,000-word guide; you shipped a 600-word product page. The page never ranks above position 40. Same problem in reverse: writing 4,000-word guides for transactional money keywords where the SERP rewards short product pages.
How to avoid: Filter every cluster by Intent before writing. Informational → guide format. Commercial → comparison or listicle. Transactional → product/landing page. Match the SERP's content format, not your assumption.
Optimizing for volume only
What goes wrong: You chase the highest-volume keyword in every category. SERPs are dominated by 80+ AS competitors. You can't break top 20. Meanwhile, the long-tail KD-15 cluster behind it would have ranked top 3 in 60 days.
How to avoid: Sort filtered keyword lists by Traffic Potential, not headline Volume. Or sort by (Volume / KD) ratio. The realistic opportunity is often in the long-tail.
Not cross-checking the SERP for parasite-dominance
What goes wrong: You target 'how to lose weight' (KD 88 but high volume). The SERP is 8 of 10 results from Mayo Clinic, WebMD, Healthline. You can't compete regardless of how good your content is. 40 hours of writing wasted.
How to avoid: Always click through to the live SERP for top-priority keywords. If 5+ of the top 10 are major authority sites (.gov, .edu, .org, top-tier media), skip the keyword. Find a long-tail variant where the SERP is competitive content sites.
Setting word count by guesswork
What goes wrong: Your content team defaults to 1,500 words. Half the keywords you target reward 3,500+ word guides; the other half reward 600-word definitions. Half your content is too short to rank; the other half is bloated and bounces.
How to avoid: Use SERP Analysis in Keyword Magic Tool to read the median word count of positions 3-7. Match that target. Don't guess.
Recap
Done — what's next
How to set up a SEMrush project the right way
Read the next tutorial
Hand it off
Running keyword research once is a project. Running it monthly across 5-10 clusters, tracking which clusters convert, and feeding learnings back into the content brief is a job. EverestX SEO content specialists run end-to-end content engines — typically $400-1,000/mo at $14-16/hr for ongoing programs.
See specialist rates
Per cluster: 100-300 filtered keywords feeding 1 article. Don't research 5,000 keywords across 50 clusters before writing — you'll never publish. Research one cluster, ship it, then research the next.
Apply the funnel: filter by KD ceiling (matched to your AS), filter by Intent (pick the right one for your content type), strip questions and brand terms, sort by Volume × (100 - KD) descending. You'll be down to 200-500 actionable keywords in 15 minutes.
Keyword Overview is a single-keyword deep dive — volume, KD, intent, SERP analysis, trend. Keyword Magic Tool is the bulk discovery tool — feed it a seed and get 30,000 related keywords. Use Magic for discovery; use Overview for deep analysis of your top 5-10 priority keywords.
Mostly, but sanity-check borderline cases. Commercial-vs-Informational gets mislabeled ~10-15% of the time. For your top 20 priority keywords, click through to the live SERP and confirm the content format matches the intent label.
Magic Tool returns more raw variations (often 2-3x more keywords for the same seed). Keywords Explorer returns a smaller, cleaner set with the Parent Topic clustering already applied. Magic is better for exhaustive discovery; Keywords Explorer is faster for actionable shortlists. Both produce ranked content; pick based on workflow preference.
On established domains (AS 40+): 2-6 weeks for initial signals, 8-16 weeks for stable top 10. On new domains (AS < 30): 8-16 weeks for initial signals, 6-12 months for stable top 10. Track in Position Tracking with the cluster tag.
SEMrush
A SEMrush Project is the container that pipes data into Site Audit, Position Tracking, On-Page SEO, Listing Management, and Social Tracker. Configure it wrong on day one and every downstream module produces noise for months. This is the right setup.
SEMrush
Most SEMrush Guru subscribers pay for the Content Marketing Toolkit and use one of its six modules — Topic Research. The other five (SEO Writing Assistant, Content Audit, Post Tracking, Brand Monitoring, ImpactHero) are the difference between random content and a content engine. This is the workflow.
SEMrush
Most teams 'do competitor research' by reading competitor blogs. The real workflow surfaces the keywords competitors rank for, the backlinks they earned, and the content gaps you can win — in about 4 hours. Here's the right sequence.
Ahrefs
Keywords Explorer is the most-used Ahrefs module and the easiest to use badly. This walks through the operator workflow — intent first, Parent Topic second, raw volume last.
SEMrush
You're paying $139-$500/mo for SEMrush. The question isn't whether the tool is worth it — it's whether you have the time and skill to actually use what you're paying for. This is the honest framework.