Technical SEO platform
Know exactly what's holding your site back
Seodar crawls your site, runs 195 technical checks and shows you which fixes will have the biggest impact.
- 195 technical SEO checks, in six categories
- Ranked by the impact each fix actually returns
- Joined to Search Console, Analytics, Bing and Cloudflare
- Track what you fixed, and whether it moved
Free, no account, no card. We check up to 25 pages and respect your robots.txt.
seodar.io
- Search Console
- Analytics
- Bing
- Traffic
Latest report · 200 pages · 2 hours ago
Your site needs work.
Open the report4critical31warnings
- technical checks, in six categories
- 195technical checks, in six categories
- pages scanned free, with no account
- 25pages scanned free, with no account
- AI surfaces watched - four assistants and Google's AI answer
- 5AI surfaces watched - four assistants and Google's AI answer
- data sources joined to the crawl, on the same row
- 4data sources joined to the crawl, on the same row
- point health score, with its arithmetic shown
- 100point health score, with its arithmetic shown
Or watch it run
How it works
No account. No card. Up to 25 pages free.
01Crawl
One URL is the whole setup
We follow your internal links from there, respecting robots.txt and Crawl-delay, and identify as SeodarBot the whole way.
- Indexability
- Metadata
- Content
- Links
- Performance
195 checks, six categories
02Analyze
Every page meets 195 checks
Indexability, metadata, content, links and media, performance. Each one scored and weighted, with a cap per category.
HTTP does not redirect to HTTPS
1 page · splits your ranking signals
03Prioritize
In the order that pays
Ranked by the score each fix returns for the effort it costs - not four hundred warnings in a list for you to triage.
Each crawl kept, so a fix has a number after it
04Grow
And whether it moved
Every crawl is kept, so a fix is followed by the number that answers for it. The score states what each category cost, before and after.
A score you can argue with
One number out of 100, with the arithmetic shown: penalties per category, weighted by severity and capped so no single problem drowns the rest.
It is a rate per page, so a 5,000-page site and a 200-page one can be read against each other - a count would tell you mostly how large the crawl was.
Prioritize what matters
Stop guessing.
Start fixing.
Sorting by severity puts ten critical findings on one obscure page above a warning that affects every page you have. We rank by what each fix returns for the effort it costs.
Prioritized by impact
Impact x ease x affected pages. One broken target linked from forty pages is one finding, not forty.
Everything connected
Search Console, Analytics, Bing and Cloudflare on the same row as the crawl - so the four can disagree, which is where the work is.
Measure what you fix
Every crawl is kept and the score shows its arithmetic, so a fix is followed by the number that answers for it.
Built for power users
A read-only API, an MCP server and your own rules on every page - the data is yours to query, not only to look at.
| Issue | Impact | Ease | Pages | Priority |
|---|---|---|---|---|
| Multiple title tags | High | Easy | 1,234 | 9.6 |
| Slow LCP | High | Medium | 987 | 8.7 |
| Missing meta descriptions | Medium | Easy | 2,456 | 7.5 |
| Duplicate content | Medium | Medium | 1,104 | 6.2 |
| Low content volume | Low | Easy | 3,210 | 3.1 |
Fixing Multiple title tags on your templates would clear it from 1,234 pages at once.View issue
An illustration. The ranking is the product’s own: impact x ease, scaled by how many pages carry the finding.
And once the crawl is clean
Technical health is table stakes. The rest of the product is about whether anybody can find you - in the results, and in the answers above them.
All 14, in the three groups the product is built in. One in each shows what it hands back: an illustration, not a customer’s site - the shapes are real and the numbers are made up.
Analyse
On your own pages, and in the searches that lead to them.
5 of 14
- Keyword researchA word in, and the searches people actually make around it out - with volume, difficulty and intent.
- Technical audit195 checks, one score, and a ranked list of what to fix.
- SegmentsScore the blog and the shop separately, from one crawl.
- Custom checksYour own rules on every page - and out of the score, on purpose.
- AccessibilityWCAG 2.2 A and AA on a rendered sample - what a machine can prove.
Visibility
In the results, in the answers above them, and on other people's pages.
5 of 14
- AI visibilityWhether four assistants and Google's AI answer cite you.
- Search ConsoleSix analyses over your own Google data - and two that need a crawl too.
- Rank trackingDaily positions, SERP features, and who else holds them.
- Competitor analysisThe same crawler and the same checks on both sides.
- BacklinksWhich domains link to you, and which ones you lost this week.
Act on it
On a schedule, in a client's inbox, or wired into what you already run.
4 of 14
Four connected sources
A crawl can only see your site. It cannot see what happened to it.
Cloudflare knows which crawlers came and what they were handed. Search Console knows what Google showed and who clicked. Analytics knows what they did next. Bing holds a second index, and reports how much of your site is in it - which Google never tells you.
Every one of those is a tab you can already open. What none of them can do alone is disagree - and the disagreement is where the work is. A page that is perfect in the HTML and a 404 to Googlebot looks fine in four tools out of five.
/pricingone page, five answers- Your crawl200, indexable, canonical correct
- CloudflareGooglebot fetched it 41 times and got 404
- Search Console12 impressions, average position 18
- Analytics3 sessions, none of them engaged
- Bingnot in the index
An illustration, not a customer’s site - the shape is drawn from real findings on connected sites.
Whether the assistants recommend you
Four assistants and Google’s AI Overview, asked about your market and recorded: the answer, the sources, and which domains were named instead of yours.
A site can hold position three and be absent from every AI answer about its own category. Rank tracking will not tell you that.
Built to be checked
Every number in the product can be traced to what produced it.
The arithmetic is shown
The score states what each category cost. A number you cannot decompose is a number you have to trust.
Measurements carry their basis
Every crawl records which catalogue scored it, and two scores from different catalogues are never subtracted silently.
Sample size travels with the rate
A rate over eight pages is not a rate over five hundred, and the product says which it is.
Nothing runs unasked
Scheduled scans, rank checks and AI calls each state their cost before they run, and stop the moment you pause them.
The crawler behaves
robots.txt, Crawl-delay, a delay between requests, and an honest user agent - on your site and on anybody else's.
Read it from code
A read-only API and an MCP server, so the data is yours to query rather than only to look at.
What it will not tell you
It will not tell you that fixing a technical finding will move you up the results. Nothing can - and a tool that implies it is selling you a causal story the data does not support.
What it will tell you is exactly what is wrong, how much of your site it affects, what it costs to fix, and whether the sites ahead of you have the same problem.
Start free
One site and 100 credits - up to 200 pages a scan, about 5 scans, and the whole report. No card. Paid plans add more sites, bigger crawls and the AI rewrites.