Across common tools, a good SEO score sits near 80–90 for site audits and 90+ for Lighthouse/PageSpeed, with context shaping what “good” means.
Ask ten marketers about a “good” SEO score and you’ll hear ten ranges. That’s because the number depends on the tool, the page type, your market, and the goal. This guide lays out how the main scoring systems work, where healthy ranges tend to land, and how to use scores to make smart, measurable gains.
What An SEO Score Usually Measures
SEO tools bundle dozens of checks into a simple number. Some scores center on technical health. Others lean on page speed. A few blend content checks or link data. The scales look similar, but the inputs differ a lot. That’s why two tools can rate the same page with different numbers and both can be right for their own method.
| Score/Tool | What It Measures | Scale |
|---|---|---|
| PageSpeed/Lighthouse “Performance” | Field-like lab tests for load speed, responsiveness, layout stability | 0–100 (90–100 = green) |
| Site Audit “Health” | Technical crawl issues (status codes, canonicals, hreflang, meta, links) | 0–100% (higher is better) |
| SEO Category In Lighthouse | Presence of basics (title, meta, links, indexing signals) | 0–100 |
| Core Web Vitals Labels | Real-user thresholds for LCP, CLS, INP | Good / Needs improvement / Poor |
| Link-Based Metrics (DA/DR) | Third-party estimates of link strength | 0–100 (comparative) |
What Counts As A Solid SEO Score Today
For page speed, a solid target is a Lighthouse or PageSpeed score in the 90–100 band on key templates. Hitting that range on every URL can be tough, and that’s fine; aim for your money pages first. For crawl health, many teams treat 80–90% as a clear win on large sites, nudging upward once the biggest errors are gone. For Core Web Vitals, land in the “Good” label for each metric; that status reflects real user outcomes, not just lab runs.
These ranges are benchmarks, not hard rules. A big media site with heavy ads may live in the mid-80s on speed but still win because of content depth and demand. A boutique SaaS site might push past 95 on speed but still need links and topical breadth to climb. The right score is the one that aligns with your business goals and moves real metrics such as conversions, leads, and revenue.
Why A Single Number Can Mislead
Scores condense complex systems into a tidy badge. Handy, but easy to misread. A perfect number on a test page won’t fix thin content. A high link metric won’t fix broken canonicals. Chasing a badge while ignoring page intent or search demand turns into busywork.
Use scores as a compass, not a finish line. Track them, but pair them with outcomes like sign-ups, qualified traffic, time on page for key articles, and error counts trending down. When the numbers disagree across tools, lean on the method that best matches the change you’re shipping.
How The Popular Scales Work In Practice
PageSpeed Insights And Lighthouse
The “Performance” score blends metrics like Largest Contentful Paint, Cumulative Layout Shift, and Interaction to Next Paint. The color bands are simple: green starts at 90, orange sits at 50–89, red below 50. This score can swing run-to-run, so use multiple runs and test on real devices when you can. See Google’s own wording in the PageSpeed Insights scoring page for banding and method.
Core Web Vitals Thresholds
Core Web Vitals label pages as Good, Needs improvement, or Poor across LCP, CLS, and INP. Those labels come from field data in the Chrome User Experience Report when available. Hitting Good means users see the main content fast, layouts stay put, and taps feel snappy. Google’s web.dev article explains why LCP near 2.5s, CLS under 0.1, and INP under 200 ms map to Good. You can read the background on threshold setting in the thresholds explainer.
Technical “Health” Scores From Audits
Audit scores roll up errors and warnings from a crawler. A site with lots of soft 404s, redirect chains, duplicate titles, or noindex mix-ups will see a lower number. Fixes in one class (say, broken internal links) can jump the score faster than nibbling at many tiny issues. Treat the number like a dashboard light; the fix list is where the value lives.
Link-Based Metrics
Link metrics compare domains by link patterns. They help with competitive research and prospecting, but they’re third-party estimates. They’re handy for benchmarking against peers and spotting gaps, yet they don’t replace search demand or content quality.
How To Pick Targets Without Guesswork
Start With Templates, Not Single URLs
Group pages by template: product, category, blog post, hub, tool. Run audits and speed tests on one sample per template. Fix shared code first—headers, scripts, image handling—so gains roll out sitewide.
Compare Against Direct Competitors
Pull speed and health readings from three to five rivals on the same query set. If the top results sit at 85+ on speed for that template, push yours past that mark. If rivals pass Core Web Vitals and you don’t, that’s your first move.
Protect Real Users
Always check the field data. A lab-green score with field-red Vitals means users feel slowness somewhere. Prioritize image weight, font loading, and script bloat on pages with the worst field readings.
Score Ranges That Map To Outcomes
Here’s a quick way to read common ranges across tools. Treat them as waypoints during sprints, not absolutes.
Lighthouse Or PageSpeed “Performance”
- 90–100: Ship small tweaks; watch regressions; invest in real-user monitoring.
- 70–89: Tackle render-blocking scripts, image sizing, and third-party tags.
- 0–69: Fix blocking resources, server delay, and massive images first.
Technical Crawl “Health”
- 90–100%: Strong baseline; schedule periodic recrawls; watch for edge cases.
- 70–89%: Prioritize 404s, broken redirects, duplicate titles, thin pagination.
- 0–69%: Fix indexation and status code issues before anything else.
Core Web Vitals Labels
- Good: Keep it that way with budgets and alerts.
- Needs improvement: Set a sprint to push one metric into Good on main templates.
- Poor: Pick one metric per template; deploy focused fixes; retest in field data.
How To Raise Scores Without Churn
Cut Weight First
Compress and resize images, lazy-load below-the-fold media, remove unused CSS, and defer scripts. Limit chat, heatmaps, and tag managers on templates that need speed most.
Serve Faster HTML
Cache pages near users, enable HTTP/2 or HTTP/3, and keep TTFB low. Many wins come from server-side work that never touches your content team.
Stabilize Layouts
Reserve image and ad slots with width and height. Avoid moving buttons. Preload fonts. These small changes lift CLS and improve taps.
Fix Indexing Traps
Audit canonicals, robots directives, pagination tags, and hreflang. Clean internal links so every template links to the next step a reader needs.
Strengthen Topical Depth
Build supporting pages that answer related tasks. Link them in both directions. A strong topic hub plus clean crawl paths can move rankings even if the speed number barely changes.
What A “Good” Number Can’t Tell You
No universal score decides rankings. Google runs many systems in parallel—link signals, content understanding, user experience, and more. You can read their public ranking systems guide to see how different pieces weigh in. Treat any single badge as a health hint. Real wins come from solving searcher tasks better than rivals.
Common Edge Cases And How To Handle Them
Single-Page Apps
Route-based code splitting and server-side rendering can turn a sluggish SPA into a quick, crawlable site. Pair that with linkable URLs and plain anchors so search engines and users can share states cleanly.
Ad-Heavy Layouts
Limit network calls on first paint. Preconnect to the ad host. Cap ad density and avoid first-screen slots. Reserve space for units to prevent layout shifts.
Large Media Pages
Transcode images to WebP or AVIF. Stream video with modern players. Host thumbnails locally and defer third-party embeds until interaction.
International Sites
Set correct hreflang pairs, keep one canonical per cluster, and align language in titles, headings, and body. Slow geographies benefit from a CDN and lighter image presets.
A Practical Target Sheet You Can Use
Use the sheet below to map a score to a plan. Pick the row that matches your current state, then commit to one sprint per row.
| Score Range | Meaning | Next Steps |
|---|---|---|
| Performance 90–100 | Fast loads and stable layouts on key pages | Add budgets and monitoring; lock in guardrails |
| Performance 70–89 | Decent speed with room to trim scripts and media | Defer heavy JS; compress media; preconnect critical hosts |
| Performance 0–69 | Blocking code and heavy assets slow first paint | Kill render-blocking CSS/JS; resize images; fix server delay |
| Health 90–100% | Clean crawl, tidy metadata, few errors | Schedule monthly crawls; fix regressions fast |
| Health 70–89% | Mixed issues across templates | Prioritize 404s, redirect loops, duplicate titles, thin pages |
| Health 0–69% | Widespread indexation and status errors | Fix canonicals, robots rules, sitemaps; repair internal links |
| Vitals: Good | Field users see fast paint and responsive taps | Keep budgets; prevent regressions during launches |
| Vitals: Needs Improvement | One metric drags; users feel some lag | Pick LCP, CLS, or INP per template; ship focused fixes |
| Vitals: Poor | Users wait or see shifting layouts | Cut JS, compress media, set dimensions; re-test field data |
How To Report Progress To Stakeholders
Show Before/After On Real Pages
Grab one template that earns revenue and chart speed, Vitals, and conversions before and after. Keep the view simple. One slide per metric beats a cluttered dashboard.
Tie Scores To Dollars
Translate a Vitals win into bounce reductions and session growth on product or lead pages. Attach a dollar estimate from analytics. That story lands better than a green badge alone.
Set Guardrails
Create performance budgets by template: max JS, max image weight, target LCP, target INP. Add checks to CI or your CMS so regressions stop before launch.
Mistakes To Avoid When Chasing A Number
- Breaking UX for a perfect badge. A stripped page that no one reads doesn’t help the business.
- Ignoring field data. Lab green with field red means users still wait.
- Treating link scores as gospel. They’re useful, but not from the search engine.
- Fixing tiny warnings while big errors linger. Clear the roadblocks first.
FAQ-Style Myths, Debunked In One Line Each
“Does One Magic Score Decide Rankings?”
No. Many ranking systems run together; the public docs lay that out in plain terms on Google’s site.
“Do Link Badges From Tools Equal Rank Boosts?”
No. They’re handy for comparison and outreach planning, not a direct lever in search systems.
“Is 100 The Only Acceptable Result?”
No. Gains near the top get harder and bring smaller returns. Ship what helps users and your funnel.
Bottom Line: Use Scores To Guide, Not To Dictate
A healthy target looks like this: pass Core Web Vitals on key templates, hold a green performance badge on money pages, keep a clean crawl, and grow topical depth with clear internal links. Track those signals each sprint and tie them to real outcomes. That mix beats any single badge, every time.