[ Home / About / Contact / WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/cont/ - Content Strategy

Content marketing, copywriting & editorial calendars
Name
Email
Subject
Comment
File
Password (For file deletion.)

File: 1790085245434.jpg (349.11 KB, 1024x1024, img_1790085237591_cpwkyva7.jpg)ImgOps Exif Google Yandex

1c6cc No.2236

was reading about how one old comment can literally shape what ChatGPT tells people about your product today. siege media is apparently doing this whole thing where they map queries to specific subreddits just to get cited in ai overviews, which sounds super intense. ngl i don't know if that's even sustainable for small brands since it seems like a lot of work for zero tiny gains. turns out 18% of adults are using reddit for answers weekly now so we might actually have to care about this.

more here: https://www.siegemedia.com/strategy/best-reddit-marketing-agencies

1c6cc No.2258

>>2236
mapping queries to subreddits sounds exhausting and way too much overhead for most of us. i had a client try to flood a specific thread with positive mentions last month and it totally backfired because the users smelled the marketing vibe immediately. if people start spotting the pattern, the ai is just going to learn to ignore those dead-end threads.

1c6cc No.2267

File: 1790482080708.jpg (145.74 KB, 1024x1024, img_1790482064861_7ztikho3.jpg)ImgOps Exif Google Yandex

>>2258
>the users smelled the marketing vibe immediately

classic mistake lol. if you're just dropping brand names in threads without actually participating in the community, you're dead on arrival. it's not even about the reddit algorithm at this point, it's about not being a total creep. i've been experimenting with some low-effort seeding on niche hobby subs and the only way to stay under the radar is to be useful first. just leave a tip or something then maybe mention how a specific tool helped you. did siege media actually use any automated scrapers for that query mapping thing? curious if they're just running custom scripts or using some off-the-shelf monitoring software.

it's basically just modern astroturfing

1c6cc No.2277

>>2267
The "low-effort seeding" thing sounds risky without a much longer test window to see if it actually sticks. I'm not sure how you're even tracking the conversion lift from those specific hobby subs versus organic traffic. Did you measure any change in brand sentiment or just raw clicks? It feels like there are too many variables with the Reddit algorithm changing every other week.



[Return] [Go to top] Catalog [Post a Reply]
Delete Post [ ]
[ Home / About / Contact / WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]