Every few weeks someone asks me the same question in almost the same words: "So do I fix my pages first, or do I go get backlinks?" It usually comes from a founder or a marketing manager who has maybe ten hours a week for SEO and wants to know where those hours go. And every time, I give a version of the same answer: you're asking the wrong question, because the split between on-page and off-page SEO isn't a to-do list, it's a diagnosis.
Here's what I mean. On-page SEO is everything you control on your own domain: titles, headings, internal links, page speed, the words you actually write. Off-page SEO is everything other people say about you elsewhere: links, brand mentions, reviews, how often your name gets typed into a search bar. One is a monologue. The other is reputation. And Google, whether you like it or not, weighs both.
Key Takeaways
- On-page work is fast, cheap and entirely under your control; off-page work is slow, expensive and depends on other humans deciding you're worth citing.
- Neither one rescues the other. A brilliant page with zero external signals stalls. A pile of links pointing at a thin page just wastes the links.
- The honest split for most small sites: 70% on-page effort for the first six months, then the ratio flips.
- Technical SEO is a third bucket, not a subcategory of on-page, and conflating the two is how people miss crawl budget problems.
- Link spam penalties are real and they don't always announce themselves. A sudden drop in rankings is usually the first notification you get.
On-page vs off-page SEO: what actually separates them
The clean way to think about it: if you can change it by editing your own website, it's on-page. If it requires someone else's action, it's off-page. That's the whole taxonomy. Everything else is detail.
But that simple line hides something important, which is that the two aren't weighted equally at different stages of a site's life.
The diagnostic question nobody asks
When a page isn't ranking, most people assume they need more links. In my experience, roughly two out of three times it's the opposite: the page itself is the problem. I watched this happen on a client project in 2024 — a services page stuck at position 14 for eight months despite a reasonably healthy backlink profile. We rewrote the H1, restructured the H2s around actual search intent, added four internal links from higher-authority pages on the same site, and compressed the hero image from 1.8 MB to 90 KB. It hit position 4 in five weeks. No new links. Not one.
That's not an argument that links don't matter. It's an argument that links amplify something that already works. Point them at a broken page and you've just bought a louder version of the same failure.
Why off-page SEO feels like waiting
Off-page is miserable precisely because you can't do it alone. You can't decide to have authority. You can earn it, buy it (badly), or wait for it. And the wait is measured in quarters, not weeks.
Real talk: I once spent eleven weeks on digital PR outreach for a B2B tool. Dozens of personalised pitches. The result was two placements, one of which was a niche industry blog with maybe 4,000 monthly readers. The other one, a mid-tier publication, sent about 300 referral visits in a single day and the domain rating crept up a couple of points over the following months. That was a good outcome, and it still felt like losing for the first ten weeks.
On-page and off-page SEO examples, side by side
Abstract talk gets you nowhere. Here's how the same intent plays out in each bucket.
| What you're trying to achieve | On-page move | Off-page move |
|---|---|---|
| Signal that a page is about a topic | Reword the title tag around the phrase people actually search | Get mentioned in an article covering the same topic |
| Build topical authority | Internal links from your strongest pages to your newest | Earn links from sites that are themselves authoritative on the subject |
| Increase click-through from results | Sharpen the meta description, add structured data | Reviews, mentions and brand recognition that make people click your name |
| Survive an algorithm update | Improve the depth and accuracy of the content | Diversify your link sources so no single one can hurt you |
| Rank in local results | Location pages, NAP consistency on the site, hours and services markup | Local citations, directory listings, consistent business info across the web |
Notice how the right-hand column always involves another party. That's the structural difference. On-page is a decision. Off-page is a negotiation.
Where technical SEO fits, and why people get it wrong
Most articles fold technical SEO into on-page. I think that's a mistake, and it causes real problems. If your site is generating thousands of near-duplicate URLs through faceted navigation, no amount of title tag tweaking will save you. That's a crawling issue, not a content issue.
Practically, I treat it as three buckets:
- On-page: content, headings, meta tags, internal linking, image alt text
- Technical: crawlability, indexation, site speed, mobile rendering, structured data, canonicals
- Off-page: backlinks, brand mentions, reviews, social signals, unlinked citations
The distinction matters because the fixes live in different places and take different amounts of time. A canonical tag is a ten-minute change with a potentially huge effect. Rewriting a content cluster takes a month. Earning twenty quality links takes a year, if you're lucky.
Off-page SEO techniques that still work
Link building has an ugly reputation and mostly deserves it. But there are approaches that hold up, and they share one trait: someone has to genuinely want to reference you.
What earns links without begging
- Original data. Publish something nobody else has measured. Even a small survey of your own customers can generate citations for years.
- Free tools that solve a narrow problem. A calculator, a checker, a template. Boring, useful, endlessly linkable.
- Genuine expertise. If you're a practitioner writing about something you've actually done, journalists and bloggers will quote you. This is where the E-E-A-T signals come from, and they're increasingly how search engines decide whether to trust a page on sensitive topics.
- Local presence. For service businesses, consistent listings across directories, review platforms and maps matter far more than any guest post.
- Relationship-based outreach. Not mass email. Actual conversations with people who already know your work.
And here's the part people skip: unlinked brand mentions still count for something. If a publication names your company without linking, that's a signal. It's weaker than a followed link, but it's not nothing.
What to avoid, and why the risk is real
Link schemes, paid link networks, comment spam, private blog networks. All of it still exists and some of it still works temporarily. But penalties don't arrive with a notification. They arrive as a slow, mysterious decline in rankings that you spend three months diagnosing.
I made this mistake early on. Back in the days when I genuinely didn't know better, I bought a package of "high authority" links for a side project. The site ranked well for about four months, then fell off a cliff for its main keyword and never recovered, even after disavowing the worst offenders. I eventually rebuilt the project on a new domain. Cost of the cheap links: a few hundred dollars. Cost of the rebuild: about eight months of work.
How to measure which side is holding you back
You don't need expensive tools to figure out where your bottleneck is. You need two sources of truth.
First, search performance data: which queries you're showing up for, where you rank, what your click-through rate looks like. If you rank in the top ten for a term and nobody clicks, that's usually an on-page problem — your title and description aren't convincing. If you don't rank at all for terms you should obviously be relevant to, that's often an authority problem, which is off-page.
Second, a link profile review. Not the score. The actual referring domains. Ask three questions: are these sites real, are they topically related to me, and would I be comfortable if a competitor saw this list? If the answer to any of those is no, you have work to do before you build anything new.
A prioritisation rule I actually use
When time is limited, I work in this order:
- Fix anything blocking indexation or crawling. Nothing else matters if the page can't be found.
- Rewrite the pages that already get impressions but poor clicks. This is the fastest win available to almost anyone.
- Add internal links from your strongest pages to the ones you want to push.
- Only then start outreach, and only with something worth linking to.
In my experience, steps one through three produce results in four to eight weeks. Step four takes six months or more to show up. So if you're in a hurry, the order above is not optional.
The importance of off-page SEO in 2026
Something has shifted, and it's worth naming. Search results increasingly summarise answers rather than sending you to a page. When that happens, the sites that get cited inside those summaries tend to be the ones with established reputations — mentioned, quoted, referenced across the web.
Which means off-page signals are doing double duty now. They influence rankings, yes. But they also influence whether an AI-generated answer decides to name you at all. A page with perfect on-page optimisation and no external footprint is invisible to that layer. A page with a decent footprint and mediocre on-page copy might still get mentioned, because the model recognises the brand.
I'm not going to pretend I've fully figured out how to measure that. Nobody has. But the direction is clear enough: reputation is becoming a ranking asset in its own right, not just a tiebreaker.
What this means for small sites
If you're running a five-page local business site, don't panic about this. Your off-page work looks different. Google Business Profile, consistent citations, a steady trickle of genuine reviews, and being mentioned in local press or community pages. That's the whole game at your scale, and it's more achievable than competing for national links.
If you're a SaaS or content site, the bar is higher. You need at least a handful of links from sites that are recognisably authoritative in your niche. Not hundreds. A handful, from the right places.
So which one should you do first?
On-page. Always on-page, unless you have a specific, diagnosable authority gap that's clearly the blocker.
I'll die on this hill: the number of people who need more backlinks is far smaller than the number of people who believe they need more backlinks. Most underperforming sites have pages that don't answer the question the searcher asked, or don't answer it clearly, or bury the answer three scrolls down. Fix that first. Then go earn the links that make the fix count.
The uncomfortable part is that on-page work is unglamorous. Nobody congratulates you for rewriting a title tag. The link you landed from a respected publication — that gets screenshotted and shared internally. But the title tag rewrite often moves more traffic.
Which is worth sitting with for a moment. If you stripped away every external signal tomorrow and kept only what's on your own domain, would your site still be worth visiting? If the answer is yes, off-page effort compounds. If it's no, you're renting attention you can't keep.
