90 Seconds · June 2026 · SEO Content
City Blog Programme.
28 city posts. Informational intent. Capturing search queries without touching the transactional pages.
28
city posts, live
2
intent types, deliberately separated
0
cannibalisation of location pages
The problem.
90 Seconds had strong transactional location pages. Commercial pages targeting buyers searching for video production in specific cities. They were performing. The risk with adding blog content targeting the same cities was simple: if we were not careful, we would split our own authority and dilute what was already working.
At the same time, there was a real opportunity. People searching "how much does video production cost in Singapore" or "best locations to film in Sydney" are not yet in buying mode. But they are our audience. Leaving that traffic on the table was a waste.
The approach.
Intent separation. Every city post was written to informational search intent, answering the research questions a buyer has before they are ready to engage. None of them targeted the same transactional keywords as the location pages.
I authored all 28 posts end-to-end, along with the supporting infrastructure: Cloudinary hero images for each city, retitled to reflect informational framing, meta descriptions written to match the same intent distinction. Each post was published under the blog route, keeping the URL structure clean and separate from the commercial pages.
The anti-cannibalisation logic was the core design decision. It sounds obvious in hindsight. In practice, most teams either skip the city blog opportunity entirely (leaving traffic behind) or do it without the intent layer and wonder why the location pages start declining.
The outcome.
28 posts live across priority cities including Sydney, Melbourne, Singapore, Hong Kong, London, New York and more. Each with a Cloudinary hero, correct intent-separated title and meta description.
The location pages retained their positions. The blog programme opened up a new surface for informational queries without touching what was already performing. Two content types, one strategy, no cannibalisation.