Public-account analysis · Health content · Russia

1.07M followers: one feed, two different growth engines

Reach posts ask viewers to protect someone close. Capture posts exchange a useful file or video for a keyword. Across 979 videos, the second mechanic makes comments 32× denser.

1.07M Instagram followers 1,072,325 on Jul 29
8.57M Top sampled Reel 63,641 likes
194M Plays in the sample 979 public videos
1.76M Comments in the sample 1.68M on keyword posts

This is an independent analysis of a public account, not an AIAVATAR client result. Metrics come from an Apify snapshot captured July 29, 2026.

The account

Authority is stated once; every Reel converts it into a decision

The profile names Anastasia Shagarova as a doctor and promises “30 more years of life” and “20 fewer years of age.” The bio rejects operations and organ removal, supports “safe treatment of women’s health,” and sends all profile traffic directly to Telegram.

The content rarely spends time proving credentials inside the hook. It starts with a familiar object or symptom: chicken, a mug, a child’s shoe, a tea bag, a belly. Medical authority becomes the frame around an immediate domestic choice.

The July 29 snapshot shows 1,072,325 Instagram followers, 1,437 public posts and a Telegram link. Telegram’s public preview displayed 109,043 subscribers during research. That is a destination count, not proof that every subscriber came from Instagram.

  • Account: @dr.shagarova
  • Research sample: 991 posts, including 979 videos
  • Period in sample: Jan 2024–Jul 2026
Content system

Four fixed choices make hundreds of topics feel like one feed

One authority figure 01

The same doctor appears in a white coat, tailored suit or practical studio setting. Wardrobe changes; status does not.

One visible object 02

Food, shoes, cups and body diagrams explain the topic before sound is required.

One emotional stake 03

The consequence is attached to someone important: a child, mother, partner, friend or the viewer’s future self.

One next action 04

A post asks for a share, save, follow or keyword. The call to action matches the job of that particular Reel.

Anatomy of the 8.57M Reel

A grocery object becomes a family-health warning

The winner does not depend on a new medical discovery. It packages a familiar purchase as an unresolved risk.

  1. 01
    Put the evidence object in the first frame

    The doctor holds chicken close to camera. The topic is legible before the viewer processes a sentence.

  2. 02
    Turn a purchase into a hidden process

    Industrial footage and diagrams widen the story from one product to the entire production chain.

  3. 03
    Move from mechanism to personal consequence

    The script links the product to the viewer’s body and family. The information is no longer abstract.

  4. 04
    Resolve the tension with a social action

    “Send this to someone you care about” lets the viewer act without buying or researching anything.

The core insight

Reach and capture are deliberately separated

The account does not force one call to action onto every video. Broad warnings use care as distribution: share this with your mother, partner, friend or another parent. The viewer becomes the media buyer because forwarding the Reel signals concern.

Keyword posts have a different job. They name a specific symptom or desired outcome and promise a file, recipe, exercise set, video or live-session recording. “Вздутие,” “Пузырь,” “ЖКТ,” “Холестерин,” “Капуста” and “Живот” are not decorative engagement prompts; they sort demand by problem.

The difference is measurable. In the 979-video sample, 489 keyword posts generated 1,677,448 comments from 74.9M plays: 2.24 comments per 100 plays. The 290 share-only posts generated 64,310 comments from 93.1M plays: 0.07 per 100. Keyword posts make the comment signal about 32 times denser.

Two jobs

The player shows the difference better than a funnel diagram

489 / 979
Keyword videos in the sample
2.24
Keyword: comments per 100 plays
0.07
Share-only: comments per 100 plays
Keyword signal is ~32× denser
Top of the sample

The largest posts cover different life situations

Public counts captured July 29, 2026. Links open the original Instagram posts.

Portfolio, not lottery

The account is not carried by one viral post

The top three videos account for only 8.9% of all plays in the 979-video sample. That is the opposite of a cold-start account where one breakout contributes most of the headline number.

The advantage comes from a large catalogue of repeatable situations: food warnings, household habits, women’s health, digestion, posture, children and simple exercise. Each topic can be rewrapped with a new object, consequence and call to action.

Volume is not a substitute for selection. It is the mechanism that produces enough observations to find durable combinations — then repeat them without making every upload identical.

The funnel

A comment is both delivery request and demand label

Reach layer
  • Familiar object or symptom
  • Risk framed through someone close
  • Share, save or follow
Capture layer
  • Specific problem named
  • Keyword left in comments
  • Relevant file or video delivered
  • Profile link continues into Telegram
What to reproduce

Copy the operating system, not the medical claims

  1. 01
    Start from a high-frequency decision

    Choose something the audience already encounters weekly: a purchase, meal, symptom, routine or child-care choice.

  2. 02
    Make the object visible immediately

    A concrete object lowers comprehension cost and gives editors usable visual variation.

  3. 03
    Assign one job to the post

    Use a share CTA for broad reach, a save CTA for reference content, and a keyword only when a real matching asset exists.

  4. 04
    Name keyword assets by problem

    A viewer who writes “ЖКТ” or “Холестерин” identifies the need more clearly than someone who writes a generic “guide.”

  5. 05
    Measure reach and intent separately

    A share post can succeed with few comments. A keyword post can outperform commercially while reaching fewer people.

Risk boundary

Health virality increases the cost of being wrong

The account uses strong claims about treatment, organs, medication, food, toxins and longevity. This breakdown documents the content and funnel mechanics; it does not validate those medical claims.

For an AI-assisted health account, every factual claim needs a source, a review boundary and language that distinguishes education from diagnosis. Fear is a powerful hook; unsupported fear is also a trust and compliance liability.

The practical lesson One feed can run two systems at once: care creates distribution; a named problem creates intent.

Continue with a proven mechanism

Want a system like this for your project?

Send us your account — we’ll come back with the growth points on your own content. No call required.

Get a free account audit