In 2009 I was an intern at an internet-marketing company in Ortigas. My manager quit after one week and dumped the entire SEO workload on me — an intern who’d never even heard the term. So I googled "What is SEO?" and started blogging every answer I found, testing each tactic on my own site. That blog became SEO Hacker. The name is the philosophy: a hacker doesn’t wait for the manual, he takes the thing apart, learns how it works, and shares what he finds.

I closed my first real client in 2010 with a three-page proposal and no portfolio, at double what my job at HP was paying me. Five months later I quit HP to go all in, registered the company in 2011, and taught my first two hires everything I knew around a single table. Fifteen years later we’re one of the most established SEO firms in the Philippines — a market where most agencies don’t survive five, and where some of our clients have stayed with us for nearly all fifteen.

What makes us genuinely different is three things. First, we teach our playbook in the open instead of hiding it — that’s the "hacker" ethos, and our own blog outranks the agencies that won’t share theirs. Second, we’re already past Google-only SEO and into AI search: 30–40% of buyers now ask ChatGPT, Perplexity, or Gemini before they open Google, and we productized Answer Engine Optimization before almost anyone in the country. I coded our own Share of Voice tool in two weeks, so in the first meeting we show you your real AI-visibility number against your competitors, live — no other agency here can put that on the screen. Third, we only do white-hat, long-term, system-based work. No shortcuts, no one-off gigs. That’s why the right clients stay.

Right now, most of it comes from my website and then from my personal connections. That’s the honest answer. New business today comes through our website through the keywords we rank for and then my personal brand and the stage — the talks, the panels, the podcasts, the rooms I get invited into — plus referrals from existing clients and my network, and expansion inside accounts we already have (an SEO client adding AEO, a renewal that grows into a bigger scope). Our sharpest closing moment is the live Share of Voice demo: we run a prospect’s brand against their competitors across ChatGPT, Perplexity, and Gemini in real time and show them their actual AI-visibility number. It converts almost every time.

Where I wish it came from: a system that closes without me in the room. I want our sales team running that same SoV demo and closing AEO deals on their own, inbound flowing from our content and from Rankseer — our own SaaS tool — instead of from my calendar, and account growth happening on a schedule through structured client reviews rather than whenever I personally get pulled in. The brand and the network are a real advantage; most agencies would pay millions for the inbound a founder’s reputation drives. But a company with one critical node breaks the day that node gets sick, grieves, or takes a retreat. So the goal for the next few quarters is simple: make new business a machine the company owns, not a founder it depends on.

The thing I’ve handed over is building software. I’m not a trained developer, but in the last months I’ve built and shipped an entire SaaS platform — Rankseer (rankseer.com) — mostly through AI coding agents. Our Share of Voice tool, the one we demo live in sales, took me about two weeks; a dev team would have quoted eight months – maybe more.

The thing I went back to doing the human way is client content. We tried letting AI write client blogs end to end. Then we audited 30 days of output and found 85% of them — 96 out of 114 — had fabricated specifics: invented statistics, made-up numbers a client could get burned on. And it doesn’t rank well in Google. We shut it off that same week. Now AI does the draft, but a human verifies every fact and does the final polish, and nothing goes to a client without that step. The tool gets us 80% of the way; the last 20% is still human, and it’s non-negotiable. Same rule on relationships — every account still has a human manager and editor on purpose.

Yes, and it’s accelerating. The biggest change is that prospects show up already educated. They’ve asked ChatGPT or Perplexity about SEO and AI search before they ever become a lead, so the conversation starts further down the funnel — they’re not asking "what is SEO," they’re asking "why aren’t we showing up in the AI answers." Roughly 30–40% of buyers now go to AI before Google for a serious purchase, and for luxury and enterprise buyers – it’s effectively all of them.

The pattern we watch is: someone sees a brand mentioned in an AI answer, Googles the brand name, then converts. AI has become the discovery-and-interest layer; the website is just the closing step.

Are people arriving through ChatGPT specifically? We’re starting to see it. We track our own Share of Voice, and SEO Hacker shows up in about 29% of the AI answers in our category — around 10 points ahead of our nearest competitor — so we know we’re in the answer set. The honest caveat: clean attribution is hard. There’s no Search Console for AI telling you "this lead came from ChatGPT." That measurement gap is exactly the problem we built our SoV tool to solve.

What to actually do:

Start measuring your Share of Voice across ChatGPT, Perplexity, and Gemini this quarter. You can’t fix what you can’t see, and most agencies are flying blind on AI visibility right now. Go entity-based, not keyword-based. Build one coherent brand story across your site, your schema, and every third-party mention. In the AI era plain brand mentions carry weight even without a backlink — coherence beats any single signal. Cover all three platforms. No AI engine has won the way Google won search. Don’t bet the strategy on one. Fix the technical foundation first — schema, speed, on-page. AEO has nothing clean to point AI at if the base is broken. Use AI hard internally — build tools, draft content, cut cost and cycle time — but keep a human on the last mile and on the client relationship. Draft with AI, verify and ship with people. Move now. There’s maybe a 12–18 month window before the category fills with funded players and before Google folds a lot of this into its own stack.

What to ignore:

Don’t try to measure AEO like it’s Search Console. There’s no per-keyword click data, lift takes about six months, and chasing perfect attribution will paralyze you. Accept the fuzziness and act anyway. Don’t fully automate client content. We did — 85% fabrication rate. AI writing unsupervised will invent facts that torch client trust. Don’t fall for "SEO is dead." Traditional SEO is the foundation AEO stands on. The fundamentals didn’t die; they got a new layer on top. Don’t believe the tool is the moat. Everyone has the same Claude, GPT, and Gemini. Tone is in your fingers — the edge is the discipline of how you use them, not access to them. Ignore the shiny-object hype. Don’t wait for perfect. If you ship perfect, you shipped too late.