# ANOMALY Knowledge · Posting + Going Viral

Source: Course 4 of the ANOMALY curriculum (4 lessons, Accounts, Posting System, Split Testing, Reading Your Numbers). This is the factual reference for setting up accounts, running the daily posting rhythm, squeezing winners with split tests, and reading the numbers to decide when to build, remake, or kill.

## Setting Up Accounts

- **Email:** Gmail first; switch to Outlook once you run out of Gmail accounts.
- **Platforms:** make a TikTok, an Instagram, and a Facebook account, plus a Facebook Page.
- **One account per product** to start.
- **Username:** match it to the product. Use a personal-brand name for handmade products, a product/brand name for others. Ask Claude for a handle if stuck.
- **Logo + bio:** generate a logo with GPT Image 2 (ask Claude for the prompt). Keep the bio simple, just a name.
- **Link placement:** on Instagram and Facebook, put the link in the bio with "get yours here." On TikTok you can't add a link until you hit 1,000 followers, so use "get yours here" as the CTA until then.
- **Facebook targeting:** Facebook sometimes needs country restriction from the start, Settings → restrict to countries that actually buy (US, UK, Canada, Australia), otherwise reach can skew to countries with no purchase intent.

### Warming up a new account (before posting anything)

- Act like the ideal viewer: go to the search bar and search what they'd search (e.g., a dog product → search dog videos).
- Like, favorite, and comment on 10–20 videos, spending 5–10 minutes per platform.
- Do this warm-up on all three platforms.
- Let the account sit 30 minutes to an hour before posting the first video.
- Use that waiting window productively: make videos. You want 5–10 videos ready in the tank before dropping a product.

## The Posting System

- **Volume:** post 3–5 videos a day, native to each platform. Not 10, not 20, 3–5 is the sweet spot.
- **Spacing:** wait 2–3 hours between posts. Dumping everything at once smothers your own reach.
- **Never delete a video.** Deleting reads as bot behavior to the platform, and a flop can wake up and start climbing later on its own.
- **Captions/hashtags barely matter.** No magic formula, use ~5 loose hashtags, a mix of niche and broad, similar to what a competitor in the niche is doing. It won't make or break a video.
- **Sequencing:** your first 5–10 videos (posted before dropping a product) should be fresh concepts recreated from competitors. After that initial run, a growing share of posts become split tests of whatever is working.

## Split Testing (Squeezing a Winner)

Once a video is 5–10 videos deep and one of them pops, don't move on to a new concept, squeeze it. Change ONE variable at a time and repost:

- **Keep the hook and the song**, that's usually why it went viral in the first place; don't touch either.
- **Flip/mirror the first clip.** Same clip, mirrored, the algorithm reads it as fresh content.
- **Rearrange the build order.** Shuffle the middle build clips slightly.
- **Tweak the text.** Change one word or one emoji in the on-screen hook text.
- **Reuse clips, don't regenerate.** Build clips and end-result clips carry across videos, reuse them rather than making everything from scratch.

Why it works: fresh concepts from scratch are for the first videos only. Once something is proven, split tests let you pump out variations and keep riding the wave instead of guessing at a new concept every time, one winner can become ten variations, all riding the same proven hook.

## Reading Your Numbers (Build, Remake, or Kill)

**When to build the website:**
- Post the 3–5 videos on day one and watch all accounts, TikTok usually goes viral first.
- The signal to build: any platform crosses 5–10k views and is still climbing, PLUS you're getting demand comments (people asking where to buy).
- Build the store at that point, not before, never build the website at the start of a test.

**When to make a fresh account (targeting problem):**
- Symptom: content is good but stuck at ~300–500 views, and the audience skews to the wrong country.
- Facebook fix: Settings → country restrictions → limit to countries that actually buy (US, UK, Canada, Australia).
- If that doesn't fix it, make a fresh Instagram or Facebook account and repost your best videos there.
- TikTok shadowbans are rare, but if it happens, the fix is the same: new account.

**When to kill a product:**
- Test 5–10 videos per product. If you're around 15 videos in with no sales and nothing moving, drop it.
- The whole test should take about 3 days.
- Don't get attached, the swipe file always has the next candidate, so dropping a dead product costs nothing but ego.

**Diagnosing why a video flopped** (work through these in order):
- **The hook**, would you stop scrolling? Most flops die here. Compare honestly to the competitor's hook and figure out why theirs is stronger.
- **Watch time**, viewers leave early because the video is too slow or runs too long. Tighten the edit.
- **The sound**, wrong or dead audio. Use the sound currently running in your niche.
- **Targeting**, wrong country. Views from people who can't buy are worth nothing; fix with the account moves above.
- **The account**, capped. If every video plateaus at 200–300 views regardless of quality, remake the account (unless the content itself is genuinely weak, that's on the creator, not the account).
- **The video itself**, sometimes it's just a weaker version of the competitor's. Raise the quality and make it more engaging.

Bottom line: trust the numbers, not the feeling about a video.