How I got here
I taught science at Sikshyalaya Nepal, planning a year of lower-secondary lessons and testing whether any of it landed. Then close to six years at Goodwish Engineering as a robotics trainer, building a full training syllabus and teaching students from grade 4 up to bachelor's level, while managing the schools we ran programs with.
Writing that syllabus is where this all started, though I didn't know it at the time. Sequencing a year of robotics for a nine-year-old and for an engineering undergraduate is the same discipline as structuring a topic cluster: decide what the reader already has, decide the one thing this piece adds, put it in an order that doesn't collapse.
In 2023 I moved into content properly at Krezona Technology as a content writer and SEO. Finance and education, in a Nepali market where the obvious keywords were already held by banks and news sites. Over two years I wrote 300+ SEO articles, did the keyword research behind them, and handled on-page, off-page, technical and local SEO. Two of them still sit at #1 for terms I'm quietly proud of: "top banks in Nepal" and "highest salaries of banking CEOs".
Volume taught me the limits of volume. Writing several hundred articles is the fastest possible education in which ones were never going to work, so I moved deeper into search rather than further into production: an on-page SEO role at Yoddha Lab, optimizing pages that already existed. Auditing against intent, rewriting, running keyword-landscape analysis in SEMrush and Ubersuggest, reporting weekly from GA4 and Search Console, and holding 2-3 content projects in parallel. Targeted pages improved 25-40%.
Teaching to writing to SEO reads like a detour. It isn't. Six years in front of people who visibly stop listening the moment you're unclear is the best editing training I could have had, and it's why I distrust content that sounds authoritative without saying anything.
What I actually believe about SEO
Most content problems are decision problems. Someone publishes forty posts without deciding which reader each one serves, so all forty compete and none of them wins. The fix is usually publishing less, on purpose.
The SERP is the brief. What already ranks tells you what Google has concluded the query wants: a comparison table, a how-to, a tool, a definition. You can disagree with it, but it's expensive.
Improve before you add. An existing page has history, impressions and sometimes links. A new page starts at zero. I look hard for a page worth fixing before recommending a new one, which is not the advice that maximizes my invoice.
If I can't show you what changed, the work isn't finished. Clicks, impressions, position, top-10 counts, all measured against the same window before the work. Including the months where nothing moved, because those are the informative ones.
