How a watch marketplace publishes blog posts from a single photo
Artjom sends a watch photo to Ooih in Telegram. 87 minutes later: SEO blog article, Telegram caption, cover image — all published. No CMS, no copywriter, no designer.
How a watch marketplace publishes blog posts from a single photo
Artjom runs ChronoShop, a watch marketplace. His content workflow used to look like this: take a product photo, write a blog article, create a shorter version for Telegram, design a cover image, upload everything to the blog CMS, format the Telegram post, schedule it, publish. Three to four hours per post, minimum.
Now it looks like this: send a photo to Ooih in Telegram, say "let's do a blog post," and wait. By the time he finishes his next task, the article is live on the blog and the Telegram channel is updated.

What happened
At 16:03 on a Tuesday, Artjom sent a photo of an Audemars Piguet Royal Oak to the Ooih chat. His message was six words in Russian: "Давай пост для блога" — "Let's do a blog post."
By 17:30 — 87 minutes later — this is what existed:
- A 3-4 paragraph SEO-optimized blog article about the Royal Oak, written in the voice and style ChronoShop uses across its content
- A shorter Telegram caption adapted for the channel's audience — same information, different format, different length
- A cover image generated from the product photo, formatted for the blog header
- The article published to the blog and the caption posted to the Telegram channel
Artjom didn't write a word. He didn't open a CMS. He didn't resize an image or format a post. He sent one photo and one sentence.
The workflow Ooih learned
This wasn't the first post. Over the first few days, Artjom and Ooih developed a rhythm — and Ooih learned the rules.
Early on, Ooih published a post without showing Artjom the draft first. He wasn't happy. He wanted to approve content before it went live. Ooih absorbed that feedback immediately. From that point forward, every post followed this flow:
- Artjom sends a photo and a short prompt ("blog post," "new arrival," or just the watch name)
- Ooih drafts the article in chat — Artjom can read it, suggest changes, or just say "ok"
- Artjom approves (usually just "ok" or a thumbs up)
- Ooih publishes — blog article goes live, Telegram channel gets the caption, cover image is uploaded

No repeated instructions were needed. Ooih learned the "approve before publish" rule once and applied it to every post after that. It also learned ChronoShop's writing style, the typical article structure, which details matter for watch collectors, and how the Telegram captions should differ from the blog articles.
Why this matters for content teams
Most content automation tools do one thing: they generate text. You still have to review it, format it, create the images, upload everything to the right platform, and manage the publishing schedule. The "AI writes it" part is maybe 20% of the actual work. The other 80% is operations.
Ooih handles the entire pipeline. It generates the content, adapts it for each channel, creates the visuals, and publishes — all from one message in Telegram. The human stays in the approval loop (if they want to) but doesn't do any of the mechanical work.
For a small marketplace like ChronoShop, this changes the math completely:
- Before Ooih: 1 blog post = 3-4 hours of work. Publishing 2-3 times per week was the maximum.
- With Ooih: 1 blog post = 1 photo + 1 message + 1 approval. Publishing daily becomes realistic.

The content quality doesn't drop because Ooih has learned the style. And the consistency improves because there's no "I didn't have time to post this week" — the barrier to publishing is now a single message.
What makes this different from ChatGPT
You can ask ChatGPT to write a blog post about a watch. It will generate text. Then you copy it, paste it into your CMS, format it, find or create an image, write a separate Telegram caption, format that, and publish both manually.
Ooih doesn't just write — it publishes. It knows where ChronoShop's blog lives. It knows the Telegram channel. It knows the image format requirements. It knows to wait for approval. It knows the writing style. And it does all of this from a single chat message.
The difference between a text generator and an operator is follow-through. ChatGPT generates drafts. Ooih ships content.

Try it for your content
If you're running a marketplace, a blog, a Telegram channel, or any content operation where you're spending hours on the mechanics of publishing — try sending one photo to Ooih and see what comes back.
One chat. Every task. No setup. $99/mo flat — start in Telegram.