On-Page SEO Specialist Job Description (Hire vs Automate)

TL;DR: Most founders don't need to hire an on-page SEO specialist. They need to decide between three options: hire ($55k–$95k/yr US), retain a part-time freelancer ($800–$2,500/mo), or automate ($30–$300/mo). The threshold I use: under 50 pages or under ~$2k/mo budget, automate; over 200 pages plus a content team, hire. Everything in between is freelancer territory. This article gives you the numbers to make that call, plus the job description and interview rubric if you decide to hire.

Updated May 2026

Looking for on-page SEO automation instead of hiring? See Automated On-Page SEO.

I've hired three on-page SEO specialists since starting SEOJuice in 2022. One worked out, one was a slow goodbye over six months, and one I let go in week six because they couldn't articulate why a page wasn't ranking beyond "low DA." That hire taught me more about writing job descriptions than the two that worked. So I want to start with the question almost every job-description article skips: should you even hire this role, or should you automate it, or should you outsource it to a part-time freelancer?

Every other "on-page SEO specialist job description" piece I read while researching this (Indeed, LHH, Workable, the usual suspects) gives you a clean responsibilities-and-skills template and skips the prior question entirely. That's the question that costs founders the most money. So I'm going to take a position on it, give you the numbers, and only then hand you the template.

The Hire-vs-Automate-vs-Freelance Math

Here's the framework I actually use when a founder DMs me asking whether they should hire an on-page SEO specialist. It's not glamorous, it's an arithmetic problem.

Regional context, since I get this question from outside the US often: in the EU (DE/NL/UK), expect roughly €42k–€72k for the same hire (a touch lower in southern Europe). In LATAM (Argentina, Brazil, Mexico) you're typically looking at $18k–$38k for senior remote on-page specialists I've personally interviewed (and the talent at the top of that range is the best value on this whole table, in my honest opinion). In SEA (Philippines, Vietnam) the spread is $14k–$30k for similar levels.

The threshold I keep coming back to: under 50 pages or under ~$2k/mo SEO budget, automate. The reason is mechanical. An on-page specialist's day-one to-do list on a small site is meta titles, meta descriptions, alt text, schema, internal links, broken-link cleanup. (Of the first 12,000 sites we ran through the SEOJuice crawler, the four issues that appeared on more than half of them were all in that list. More on that below.) Those tasks compound badly at small scale, because by the time you've documented and delegated them, a human could have just done them.

Over 200 pages plus an active content cadence (more than four new pieces a month), the math flips. A specialist starts catching site-wide patterns automation can't easily see (a navigation-link template silently sending crawl budget to a 410-d archive, a recurring intent mismatch on category pages), and the hours saved across the team start to justify the loaded cost.

The mistake I made on my first hire? I was at maybe 80 pages and treated it like a 300-page problem. Wasted nine months of someone's career and ~$45k of my runway. The lesson: don't hire seniority you don't yet have problems for.

What I Actually Found in 12,000 Crawls

Because every article on this topic asserts that on-page specialists "fix the things you overlook" without saying which things, here's the actual data from sites we've audited through SEOJuice (pulled from the crawler in early May 2026, ~12,400 sites that ran a full audit at least once):

Notice the column on the right. That's the test for whether the role is worth hiring for, on your site. If most of what your audit surfaces is the alt-text / title-truncation / templated-duplicate category, you have an automation problem. If it's the broken-internal-links / thin-content / template-issue category, you have a judgment problem, and judgment doesn't compress into a SaaS subscription.

An aside on how the dataset is biased

(I want to be honest about this because I just spent two tables asking you to trust the numbers.) The 12,400 sites in that sample are self-selected: they ran an SEO audit, which means they were already worried enough about their on-page health to try a tool. So the failure rates are almost certainly higher than what you'd find on a randomly-sampled site, and probably lower than what you'd find on a randomly-sampled site that has never been audited. I'd take the directional rank order of these issues much more seriously than the exact percentages.

The Job Description (If You've Decided to Hire)

OK. You did the math, you have more than 200 pages, you're shipping content weekly, and automation isn't covering it. Here's the job description I'd use today, stripped of the parts I now think are filler.

Title: On-Page SEO Specialist

Location: Remote (or your city, if you have a real reason for it)

Reports to: Head of Content, Head of Marketing, or directly to the founder if the team is under 10.

The role in one sentence: You own the on-page health of [company]'s ~[N] indexable pages. You make sure search engines and the humans they send understand what each page is for.

What you'll actually do (week one through month three):

• Run a full on-page audit in week one. Identify the top 20 pages by impressions and rewrite their title tags and meta descriptions where data supports it.
• Build a meta-title and meta-description style guide for the content team to follow on new pages. (One of my biggest wasted hires never did this. Everything they fixed got broken again two weeks later by the content team.)
• Run a monthly internal-linking review. Identify pages with strong external authority but weak internal link distribution, and fix it.
• Own schema markup for the page types that matter: Article, Product, FAQ, Breadcrumb, LocalBusiness if applicable.
• Maintain a single dashboard showing impressions, average position, and CTR for the top 100 pages, with a monthly delta you can defend in a meeting.

What I'm not going to ask you to do:

• Link building. That's a separate role with a separate skill set, and asking one person to do both is how you get neither.
• Write content from scratch. You'll edit and brief, not draft. (Unless you want to, and we can rewrite the role.)
• Manage paid search. I see this glued onto SEO job descriptions all the time and it makes the hire two-thirds worse at both.

What "good" looks like at month six:

• Average position for the tracked top-100 pages has moved up by at least 4 positions.
• CTR on rewritten pages is up at least 15% vs. their pre-rewrite baseline (Google Search Console, same-quarter comparison).
• You can answer "why isn't this page ranking" with three hypotheses ranked by likelihood, not one canned answer.

Compensation: $65,000–$85,000 base for US/EU senior, $30,000–$55,000 for LATAM senior, full-remote, plus the usual.

That's a job description that says something. Compare it to what's on Indeed (which gives you nine bullet points of "optimize on-page SEO elements") and you'll see what I mean.

The Interview Rubric I Actually Use

The interview is where most founders lose the hire. They ask "what tools do you use?" (everyone says Ahrefs and Search Console), the candidate gives a smooth answer, the founder confuses tool fluency with skill, and six months later we're rewriting the job description. So here's the five-question rubric I run now, with what I'm actually listening for.

Question 1: Walk me through the last on-page change you made that moved a real number. What was the page, what did you change, and how did you measure it?

What I want to hear: a specific URL or vertical, a specific change (meta title, internal link structure, content rewrite), a specific metric, and a number that wasn't "rankings improved" without saying from where to where. The strong answer names the metric, the baseline, the delta, and the time window. The weak answer talks about "best practices."

Red flag: answers that float at the level of "we improved organic traffic by 40%" without anchoring to a page, a date, or a metric. (One candidate told me they'd grown organic traffic 300% at their last job. I asked which page contributed most. They couldn't name one. I should have ended the interview right there. I didn't, and that was the six-week hire.)

Question 2: Here's a page that ranks #11 for its target keyword. What are the first three things you'd check?

What I want to hear: a ranked list, ideally starting with intent match (is this the right type of page for the query?), then a content-vs-competitor comparison (what do positions 1–5 have that we don't?), then technical or internal-link hypotheses (is the page orphaned? is it canonicalized weirdly?). The order matters; if they start with backlinks or domain authority, that's an off-page brain in an on-page chair.

Question 3: When would you tell a founder not to bother optimizing a page?

What I want to hear: a real answer with a real threshold. "If the page hasn't gotten an impression in 90 days and isn't part of the topical structure, kill it or noindex it" is what I want. "Every page has potential" is what I'm afraid I'll get, and yes, I've gotten it.

Question 4: How do you handle a CMS template change that breaks 2,000 pages of meta descriptions overnight?

What I want to hear: composure, plus a sequence. Identify the affected URL pattern. Pull a sample from Google Search Console to confirm the scale. Push a fix at the template level (not page-by-page). Submit the affected URLs for re-crawling. The exact tactic matters less than the order of operations.

Question 5: What's a thing you used to believe about on-page SEO that you no longer believe?

What I want to hear: anything that shows the candidate has updated their priors. The specific answer is less important than the existence of one. ("I used to think exact-match keyword density mattered." "I used to brief writers on a word-count target instead of a depth-of-coverage target." "I used to ignore image SEO because it felt low-leverage; one client convinced me otherwise.") If they have no answer, they haven't been paying attention to their own results.

The three red flags that are dealbreakers for me:

1. They can't name a single recent Google update by name. March 2024 Core Update, the August 2024 HCU follow-up, the November 2025 helpful content tweak, something. Not because the trivia matters, but because not following them is a tell that the candidate gets their information from one-step-removed sources, not from running real sites.
2. Their portfolio metrics are all up-and-to-the-right without a single "this didn't work" story. Everyone who has run real on-page work has dropped a ranking from a meta-description rewrite that misread intent. If they haven't, they haven't done the volume.
3. They confuse on-page SEO with content writing. If "I'm a great writer" is in the top two reasons they want this role, you're hiring the wrong person. On-page work is editing and structural; if you want a writer, write a writer's job description.

The 2026 Wrinkle: AI Overviews and the AEO Shift

One thing I would not have written in the 2024 version of this article: the role itself is changing, fast. Google's AI Overviews now appear on a meaningful share of informational queries (the latest numbers I've seen put it somewhere between 30% and 50% of commercial-investigation queries, depending whose telemetry you trust; Google's own announcements have been deliberately vague about volume). What this means practically: the on-page specialist's job now includes optimizing for being cited in an AI Overview, not just ranking #1 below it.

The on-page tactics that translate: clean question-and-answer headings, FAQ schema implemented correctly, structured data on factual content, and what I've started calling "extractable paragraphs"—short, self-contained, source-citable answer chunks that an LLM can lift without losing context. (If you're adding to a candidate evaluation, the modern question is "show me a page you optimized for AI Overview citation, and tell me whether you saw the citation appear." If they look puzzled, that's not a dealbreaker yet, but it's a learning curve they need to be willing to climb.)

Remember the threshold I gave you up top (under 50 pages, automate)? It still holds, but for sites that are heavy on informational content (glossaries, guides, FAQs), the AI Overview shift is making automated meta/FAQ/schema generation more valuable, not less. Which is the opposite of what you'd expect; I would have guessed in 2024 that AI search would make automation less useful, and I was wrong about that.

FAQ: Hiring an On-Page SEO Specialist

Q: What's the average salary for an on-page SEO specialist in 2026?

A: In the US, $55k–$95k base for the role, with the median for the senior tier around $72k based on the offers I've personally seen in the last twelve months. Indeed's published median is lower (~$62k) but that mixes in junior and generalist SEO roles. EU median is roughly €58k. LATAM senior remote runs $30k–$55k; the top of that range is the best price-to-quality ratio I've found.

Q: On-page specialist vs. technical SEO—what's the difference?

A: Overlap is real but the gravity is different. On-page lives in the content layer (titles, copy, internal links, schema). Technical SEO lives in the crawl-and-render layer (robots, sitemaps, server response codes, JavaScript rendering, Core Web Vitals). If you only have budget for one, hire whichever maps to your biggest current symptom. Pages that aren't ranking despite traffic potential: on-page. Pages that aren't even getting indexed: technical.

Q: Can I automate on-page SEO instead of hiring?

A: Yes, for most of the high-volume mechanical work—meta titles, meta descriptions, alt text, internal-link recommendations, schema generation. That's what SEOJuice does, and it's why I keep saying the under-50-pages threshold matters. What you can't automate is judgment: should this page exist, what's the right intent match, why is this template duplicating titles. If your audit is mostly mechanical issues, automate. If it's mostly judgment issues, hire.

Q: How long does an on-page SEO hire take to pay back?

A: Honest answer based on the three I've made: 4–6 months on a site with existing impression volume and unfixed on-page issues; 9–12 months on a site without much existing footprint. If you're under 5,000 monthly impressions and considering this hire, that's the place I'd push back hardest—you may be hiring a year too early.

Q: Should I hire freelance first and convert to full-time later?

A: Yes, almost always. A $1,500/mo three-month freelance trial gives you a real work sample and surfaces fit issues that a 45-minute interview never will. It's also how you find out whether the role survives contact with your actual site, which is the question you can't answer from a job description.

Q: Do on-page specialists need to know how to code?

A: No, but they should be comfortable reading HTML and recognizing what Schema.org JSON-LD looks like. I used to think coding was required (specifically: I'd disqualify candidates who couldn't write a basic <script> tag from scratch). I was wrong. The ones I hired who could code well were rarely better at on-page judgment, and the ones who were great at judgment often weren't coders. Look for "can read it" not "can write it."

Related reading:

• 10 On-Site SEO Elements That Influence Search Results — the actual elements your on-page hire will be working on
• Internal Linking Statistics 2026 — the data behind why I keep pushing internal links as undervalued
• Pillar Content & Topical Silos — the structure your on-page work fits into
• Automated On-Page SEO — the automation alternative if your math comes out under the threshold