Virtual Assistant SEO Services

I start by writing down exactly what the VA will touch. Not "do SEO" – that is a trap. I list the specific pages, the fields to change, and the format each field must follow. For a recent project on a local plumber's site, I had 45 service pages that needed new title tags and meta descriptions. I built a Google Sheets template with three columns: page URL, target keyword, desired title (character count enforced by a formula), and desired meta description. I gave the VA the sheet, a limited [Semrush](/semrush/) account to check search volume and trends, and a login to the CMS with editor-only permissions.

I paid the VA $14 per hour. In-house that work would have cost about $40 per hour. Over 45 pages at roughly 15 minutes each, the VA cost $157.50. In-house would have been $450. But I knew I had to budget for my own review time. I checked the first 10 pages within 24 hours. Three titles were over 60 characters – the VA had ignored the formula. I fixed the template, added a conditional formatting rule that turned a cell red if the title exceeded the limit, and asked them to redo those three. After that, zero errors. The total review time was 2 hours. Net saving: $450 - $157.50 - (2 hours of my time at, say, $100 per hour) = $92.50. Not huge, but the VA now knows the process and next batch will be cheaper.

I also use [Microsoft Clarity](/microsoft-clarity/) recordings to spot usability issues the VA can fix, like broken CTAs or missing alt text. The VA watches a few session replays and logs the fixes in a ticket. That turns a boring audit into a visual task they can complete without me explaining every HTML tag.

One decision rule I follow: never let a VA touch a page that already ranks in the top 20. I only give them low-visibility pages until I trust their output. That hedge has saved me from accidentally de-optimising a page that was earning clicks.

The biggest watch-out is mistaking execution for strategy. A VA can fill in a template. They cannot decide which keywords matter for your business model. I once had a VA run a keyword report using [Ahrefs](/ahrefs/) and they gave me 200 long-tail terms with zero search volume. They thought more keywords = better. I had to explain that we needed terms with at least 50 monthly searches and commercial intent. Now I set a clear filter before they open the tool: minimum volume, minimum word count, and a list of excluded modifiers like "free" or "how to" unless I approve them.

Another watch-out: cheap link building. Some VAs default to posting on forums or submitting to directories I have never heard of. I now include a line in their contract that any link built without my prior written approval is grounds for immediate termination. I also run a weekly backlink audit using [Screaming Frog](/screaming-frog/) (their SEO Spider) to catch new links. If I see a .xyz domain or a site with a spam score above 30 in Moz, I ask the VA to explain or I remove the link.

Communication can be slow if the VA works in a different time zone. I learned to batch my instructions into a single message at the start of their day rather than dribbling requests. That cut turnaround time from 3 days to 1 day. But if the work is urgent, I pay a premium for a VA in my time zone.

One counter-argument I hear is that a cheap VA is better than no SEO at all. I disagree. Bad meta descriptions that get rewritten three times cost more than waiting a week to find a decent VA. A single spammy link can trigger a manual action. I would rather let a page sit unoptimised than let a VA damage it.

I hired a VA who claimed to know technical SEO. They said they could run a crawl with Screaming Frog and fix 404s. They ran the crawl, but they flagged 200 URLs as broken because they included query parameters from a session ID tracker. They spent four hours creating a spreadsheet of false positives. I wasted another hour verifying. The lesson: I should have given them a pre-configured crawl with the filters already set. Now I create a saved configuration file for every tool the VA uses.

I also messed up by not checking the VA's English level. One VA wrote meta descriptions with sentences that read like machine translation. "Our plumber services fix leaky taps quickly and efficiently." Grammatically fine, but it used "services" as a noun instead of the service page name. I had to proofread every one. That killed the time saving. Now I ask for a writing sample before hiring – three meta descriptions for a mock page.

Another mistake: I assumed a VA could handle [Surfer SEO](/surferseo/) content optimisation. They had the tool, but they did not understand the difference between recommended keyword density and forced stuffing. They rewrote a page that had 2% density for "roof repair" and pushed it to 5% by repeating the phrase in every paragraph. The content read unnaturally. I reverted and wrote guidance: "Use each keyword exactly once in the first 100 words, once more in the middle, and once in the conclusion. Do not exceed three occurrences per 500 words." That rule stopped the stuffing.

Finally, I paid $8 per hour for a VA once. The quality was so bad that the rework cost $200 of my time for a $40 task. I now pay at least $15 and test with a small batch. The higher hourly rate pays for itself in fewer errors.