Paste a URL into an AI tool, get a list of what's broken. Ask for a meta description, get one in five seconds. Most of what used to make this work valuable is templated now. Here's what that leaves.
The execution was never the hard part, even before AI. Writing a meta description, fixing a title tag, generating a keyword list, that's the visible layer, and it's the layer that got commoditized first because it was always the most templated part of the job. What's underneath it is harder to fake: knowing which of the ten things on a checklist is the one actually costing an organisation money, and which nine are cosmetic. That's not a skill an AI tool demonstrates by listing problems. It's a skill that shows up in the specific cases where the obvious-looking issue and the important issue aren't the same thing.
Relationships don't scale the way content does. A tool can generate a hundred outreach emails in an afternoon. It can't get someone to actually respond, trust the person writing it, and link back organically because they wanted to, not because they were paid to. The backlink work I've done that actually moved a domain's authority came from individual conversations, not volume.
Proof beats claims, and proof takes longer to fake than a promise does. What's harder to fabricate than "I can improve your rankings" is a specific, checkable finding: this canonical tag was misdirecting authority, this staging folder was exposed, this is what changed after it was fixed. That's why every service page here leads with real, anonymised examples, not testimonials.
Before SEO, I spent nearly four years coordinating community-based sustainability work with Delterra, a McKinsey-founded initiative, working directly with village governments and waste operators across Bali. Before that, relationship banking. In between freelance and full-time SEO work, I also spent time as a linguistic reviewer evaluating AI model responses for factual accuracy, which is where a lot of the AI-citability angle on this site comes from.
None of that is a typical path into SEO. That's mostly the point: the parts of this work that transfer, building trust with people who have no reason to trust you yet, and telling the difference between something that looks like a problem and something that actually is one, weren't learned from an SEO course. Each move, banking to grassroots NGO work to evaluating AI model outputs to backlink strategy, meant getting competent in an unfamiliar domain within months, not years. That pattern is also why I keep formal AI training current: certifications from Anthropic's AI Fluency for Nonprofits and several Google AI courses, on top of learning the practical side by doing linguistic review work directly.
Curious what a specific, checkable finding on your own site would look like? I offer a free 30-minute look.
You'll walk away with one concrete finding, whether or not we end up working together. Get in touch or book a call.
Bali, Indonesia · rajnanig22@gmail.com · linkedin.com/in/widiginanjar