The First Question I Ask a Client Is What Last Year's Revenue Was.

A global hospitality group came to us and I never asked which keywords they wanted. I asked what they closed last year and what they've got to close this year. Once they answered, I could ask the questions that follow from it, whether the shortfall sits in basket size or in how far they've penetrated a given country, which markets missed their number and which ones cleared it. Until somebody answers those, you have nothing to build a search strategy on.

Almost nobody asks, and I said so on the DailySEO ID podcast this week, in an episode built around one argument, that SEO has to speak in revenue rather than rankings. On air I guessed that 99% of practitioners never open with the revenue question. Reading the proposals that circulate in this market, I think I was being kind.

What the packaged proposal leaves out

Agencies here sell search the way a food stall sells noodles. Package A gets you five backlinks, package B gets you ten, and the premium tier stretches from ten keywords a year to twenty. Somewhere in the deck the agency drops a technical audit exported straight out of Semrush or Ahrefs, a Core Web Vitals score, and a list of meta titles that run too long. None of it is wrong, it just sits nowhere near the business it's meant to serve.

Twenty keywords a year tells a founder nothing about what her company looks like in three years, and a backlink package answers nothing about whether she gets from 5 billion in revenue to 15. She signs anyway. The campaign runs, the report lands full of positions and sessions, and her board reads it hunting for a number nobody put in there.

Where ranking actually belongs

Ranking is an indicator, and I put it underneath goals and KPIs in the Goals Triangle I use with clients. My reason is that user interest moves. Even when volume for a query holds steady, the intent behind it drifts and the behaviour around it drifts faster. You need a goal you can steer by for a year, and search demand won't hold still that long. People hear me say ranking is not a goal and decide I think ranking doesn't matter. I care about it a great deal. I want it sitting where it can tell you something true.

Why search budgets here stay small

Open the media plan of any mature Indonesian brand and you'll find paid taking the largest line, organic well behind it, and inside organic more money going to media relations than to search. Search earned that spot. Two decades of practitioners couldn't show a client what the spend returned, so clients stopped believing there was a return to show.

I have friends running brand at large FMCG companies who now prefer to buy ads or hire KOLs. You can check what that leaves behind without commissioning a survey. Open the brand sites across any FMCG portfolio in this market and count how many anyone still maintains.

Pay follows the same curve. Set what a search practitioner earns here against Singapore or the US and you're comparing the first floor to the ninetieth. A Singapore company once hired me partly because I was cheap, and in year two they let slip that their previous agency had charged three times as much. I enjoyed hearing that for about five minutes, and then spent years undoing the price I'd set for myself.

AI has taken the entry-level execution work

A junior used to walk into this industry through execution, fixing meta descriptions, running crawls, cleaning up reports, and picking up the business logic somewhere along the way. That door is closing. A business can hire a junior analyst, or it can spend a couple of hundred dollars a month on a Claude subscription and get the crawl, the analysis and the deck. Connect MCP to the CMS, Search Console and Analytics and I have a technical audit in ten minutes that I can carry into a board meeting. I won't hire a junior SEO at my own agency now, because execution is handled.

So anyone entering the field has a harder problem than picking which tools to learn. My answer runs in three steps.

• Buy a domain and ship sites in three different industries. Run your own campaigns and break your own things, because nobody gets to say they've never touched a Next.js site anymore.
• Read a campaign from the business side first. What does the company sell, who buys it, where does it sit today.
• Then walk into a real business, as an employee or a consultant, and tell them what you see in search and where you can take them.

Thirty years in search buys you nothing if you hold no working model of how an AI system picks its answer. I want to know whether someone understands how search works this year, which is why a junior with judgment beats a veteran who stopped learning in 2019. None of that saves you from bad input. An AI-assisted campaign runs as well as the practitioner driving it, and checking the output means knowing roughly what it should have said before you read it.

What people are actually typing

Traditional search ran on short queries. Someone typed cheap shoes, a few thousand sites fought over the head term, and whoever came up in that era learned to treat volume as the prize.

Watch a friend use an assistant now. She doesn't type cheap shoes. She types that she wants running shoes under a million rupiah, durable in wet weather, comfortable enough for a marathon, and available with cash on delivery. One message, the entire requirement set, and she expects an answer to all of it. She sent that message today, not in some future version of search.

A page written for the head term answers none of that. To answer it you start somewhere else, with who she is, what stops her from buying, and the exact sentences she types. We built StoryMint in that order, audience first and words second, which is why it doesn't behave like a keyword database. Put the user first and SEO follows.

The question to open with instead

The pitch doesn't need to change, only the first question does. Stop asking which keywords they want to rank for and ask what the business has to hit this year and how search can move it. Once you have that answer you can rank the technical work, the content investment and the AI visibility measurement against each other, because you know what you're optimising toward.

If your reporting still stops at positions and sessions, close that gap first. It's what we work on with clients on the measurement side. The full conversation is on the DailySEO ID podcast, SEO Harus Bicara Revenue, Bukan Cuma Ranking, about an hour, in Indonesian.