Video Scaling Strategies · Issue 8

The 30 days that decide if you get recommended

Instagram’s originality rule works on a rolling thirty day window, not a strike counter. That changes what you should be measuring, and most teams measure nothing.

In short

Instagram judges originality as a share of your recent posts, over a rolling thirty day period, not as a strike count. Meta says an account can become eligible for recommendations again when most of its recently posted reels, photos and carousels are original. It has never published where most begins, and no public source states a specific number of reposts.

Hi, I’m Lera. Every issue of Video Scaling Strategies uses what the market is actually doing to explain why some video earns attention and some doesn’t.

A long row of posts running off both edges of the frame, with a heavy window outline over four of them and an arrow showing the window sliding along.
The window slides forward with you. What counts is the mix inside it, not a tally.

There is no strike counter for unoriginal content. There is a ratio, measured over the last thirty days, and you are almost certainly not tracking it.

The rule underneath it, in Instagram’s original content guidelines: an account that mostly posts work it did not create, or did not change in a material way, may stop being recommended to people who do not already follow it. Meta then names what does not count as a material change, and the list is short and expensive.

“Examples of minor and immaterial edits include but are not limited to: borders, watermarks, speed changes, and simply crediting the original creator.”

What does not count as editing. Instagram original content guidelines, read 20 September 2026

That is a description of most repurposing workflows, which is why the rule landed so hard. But the mechanic people get wrong is not what counts as original. It is how the counting works.

This is the single most misunderstood mechanic in the originality rule, and it matters because the two models call for completely different behaviour. A strike counter says: do not do the bad thing more than N times. A rolling ratio says: your recent output as a whole decides what happens next.

Instagram describes the way back in, which is the clearest statement of how the rule works:

“If your account is not eligible to appear in recommendations, you can become eligible again when most of your recently posted reels, photos, and carousels are considered original in a 30-day period.”

How an account becomes recommendable again. Instagram original content guidelines, read 20 September 2026

Three things follow from that one sentence, and all three are operationally useful.

  1. The window rolls, so nothing is permanent Unlike a copyright strike, this does not sit on your account for a year. Thirty days of mostly original posting is the stated route back.
  2. It is a share of output, not an incident One repost in a month of original work is not the problem the rule is aimed at. A feed that is mostly other people’s clips is.
  3. Meta does not publish the threshold “Most” is the word on the page. Anyone quoting you an exact number, a ratio or a count of ten is guessing, and we are not going to guess either.

What the rule does not touch

Worth saying plainly, because the panic around this has been out of proportion. From the same page: the original content guidelines do not change how Instagram shows your content to people who already follow you. This is a rule about recommendations, which is the reach you get from people who have never heard of you.

So the symptom is specific. Followers still see you. Your reels stop reaching anyone new. If your view counts held steady while your non-follower reach fell off a cliff, this rule is a candidate, and Account Status inside the app is where Instagram says it will tell you.

Two cards. One crossed out, listing borders, captions, speed changes and credit tags. One ticked, reading a clip from footage you shot.
Origin is the whole test. Effort spent editing does not move a post between these.

Run the count this week

Twenty minutes, a spreadsheet, and your last thirty days of posts. One row per post, three columns:

ColumnWhat goes in itWhy
SourceOurs, licensed, or foundOrigin is the whole test. Effort spent editing does not move a post between these.
Change madeCut, reframe, captions, border, voiceover, commentaryMeta names borders, captions and speed changes as immaterial. Voiceover and commentary are the kind of change it points to as material.
VerdictOriginal or notAdd up the column. If it is not comfortably mostly original, you have your answer.

Most teams who run this exercise find the same thing: the unoriginal share is concentrated in filler, the posts published to hit a cadence on a thin week. Which points at the real problem, and it is not policy. It is that there was not enough of your own material ready to post.

The accounts that never have to think about this rule are the ones sitting on more footage than they can publish.

Next week: the one question that sorts every tool in your stack into safe and not safe.

Frequently asked questions

How many reposts before Instagram stops recommending me?

Meta publishes no number. Its guidance describes a rolling thirty day period in which most of your recent posts need to be original. Anyone quoting an exact count, such as ten, is not citing a Meta source.

Does the originality rule affect my existing followers?

No. Instagram's guidelines state that they do not change how your content is shown to people who already follow you. The rule affects recommendations to new audiences.

Do captions or a border make a repost original?

No. Meta names borders, watermarks, speed changes and crediting the original creator as minor and immaterial edits.

In short

The studio podcast look comes from light, lens and mic placement, not an expensive camera. A used camera with a fast lens, a gridded key light and a dynamic mic close to the mouth get you there for well under $1,000 per person. No studio? Coworking operators rent meeting rooms by the hour with no membership, from about $30 an hour in New York.

Hi, I'm Lera. Every issue of Video Scaling Strategies uses what the market is actually doing to explain why some video earns attention and some doesn't.

Three cards: about $490 for a used camera and fast lens, $210 for a key light with grid and a rim light, $80 for a USB mic and boom arm.
A studio-quality one-person setup, bought used. The camera is the biggest line and the smallest part of the look.

It's the Light, Not the Camera

When a video podcast looks expensive, the first guess is the camera. It's usually the light. Take this episode of Founder Mode (a German-language podcast for founders) with Andre Retterath, a partner at the European venture capital firm Earlybird: a warm, dim set with a wood-slat wall and a lamp, a face lit softly from one side, the background melting into blur, a broadcast mic (the large microphone radio hosts talk into) close to the mouth.

None of that needs a cinema camera (the kind used to shoot films). It needs three things: light that stays on the speaker and off the wall, a fast prime lens (a lens with one fixed zoom and a wide opening that lets in lots of light), and dynamic mics (sturdy microphones that mostly pick up what is right in front of them) close to the mouth.

An expensive camera in bad light looks cheap. A used $400 camera in good light looks like a studio.

The Camera: Two Ways to Buy It

Prices are per camera angle. Any large-sensor mirrorless body (a camera with swappable lenses and a much bigger image sensor than a phone, sold without a lens) will do. The lens, not the body, gives you the blurred background. Look for a low f-number like f/1.4 (how wide the lens opens: lower means more light and more blur) and a focal length (how zoomed in the lens is, in mm) of around 30 to 35mm.

Cheapest: buy used · about $490 per camera

Body: Sony a6000, used ($415 to $520). Lens: TTArtisan 35mm f/1.4 or 7Artisans 35mm f/1.4, new ($69 to $73). Manual focus (you turn the focus ring by hand), so set it once and sit still. It stops every 29 minutes and warms up on long takes: restart at natural breaks and power it from the wall with a dummy battery (a battery-shaped adapter that plugs into a socket). A used Panasonic G7 ($300 to $340) is cheaper still; pair it with a 25mm lens for the same framing.

Easiest: buy new · $940 to $1,170 per camera

Body: Sony ZV-E10 ($748). Lens: Viltrox 33mm f/1.4 or Sigma 30mm f/1.4 ($191 to $419). Eye-tracking autofocus (the camera finds your eyes and keeps them sharp by itself) means no one has to stand behind it. Record in HD (standard 1080p video) with the overheat setting (Sony calls it Auto Power OFF Temp) on High for takes of up to about an hour. Canon’s EOS R50 is the alternative, at up to 59:59 per clip (one continuous recording).

Lighting: Where the Money Shows

This is the part that reads as expensive. The job is to light the face and keep the back wall dark, with a key light (the main light on your face), a softbox (a fabric cover that makes that light soft and even), a rim light (a small light behind you) and accent lights (small lights on the back wall for mood).

Lighting kit for a one-person video podcast set
WhatPickPrice
Key lightGodox SL60IID, Amaran 60x S or Neewer CB60B$120 to $160
Softbox35 in (90 cm) octabox (an eight-sided softbox) with a honeycomb grid (a fabric mesh that stops light spilling sideways)$60 to $100
Rim lightSmall RGB LED (a light that can change color), like the Ulanzi VL49about $30
AccentsLED light bar, or a warm lamp you already own$0 to $50

The grid is the piece people skip. It stops the soft light spilling onto the wall, which is what lets the background fall into shadow. A rim light behind you draws a thin edge on your hair and shoulders so you don't melt into it.

Top-down set plan: the speaker sits 6 to 10 feet from a dark back wall with accent lights on it, a gridded key light at 45 degrees in front, a rim light behind to one side, the mic a hand's width from the mouth and the camera at eye level.
The whole set from above. One key light, one rim light, and accents that only touch the wall.

Audio: Dynamic Mics, Close to the Mouth

Studio podcasts sound warm and close because the mic is close. A dynamic mic a hand's width from the mouth hears the voice and very little of the room, which matters most when the room isn't yours.

Mics connect one of two ways: USB (straight into a laptop) or XLR (the three-pin cable professional audio gear uses). XLR needs an interface (a small box that turns the mic's signal into something your computer can record).

Audio kit for a video podcast
WhatPickPrice
MicrophoneFifine K688 or Samson Q2U (USB and XLR), or Rode PodMic (XLR)$55 to $100
InterfaceFocusrite Scarlett Solo for one XLR mic, Vocaster Two for two$140 to $250
MountingDesk boom arm (a jointed arm that clamps to the desk and holds the mic)from $25

The K688 and Q2U plug straight into a laptop over USB, so a solo host can skip the interface. Add one, and move to XLR, when a guest joins.

Four Placement Rules

Sit 6 to 10 feet from the back wall

Distance is what blurs it. Against the wall, even a great lens can’t.

Put the key light 45° to one side

A little above eye level.

Keep every camera at eye level

Looking up or down at a speaker reads as amateur.

Shoot a neutral, flat picture profile

(the camera’s built-in color and contrast setting). Flat keeps detail in a dark set. Use log (an extra-flat profile that looks washed out until edited) only if you will color grade (adjust the colors in editing).

No Studio? Rent a Meeting Room by the Hour

You don't need a lease or a coworking membership to record somewhere that looks professional. Most operators rent meeting rooms by the hour to anyone:

Meeting rooms bookable by the hour without a membership
WhereHow it works without a membershipNew York price
WeWork On DemandBook a room by the hour in the app and pay as you gofrom $8 a seat per hour
RegusBook meeting rooms online across 69 Manhattan locationsfrom $35 an hour
DeskpassNo membership, and it now lists the rooms of Breather (a meeting-room app it merged with in 2024)from about $30 an hour
LiquidSpaceOne free account for hourly rooms across coworking brands, podcast rooms includedfrom about $50 an hour
PeerspaceRooms and fully kitted podcast studios from local hosts (people and businesses renting out their own space), usually a 2-hour minimum$35 to $100 an hour

Then Make the Episode Earn Its Keep

A well-shot 40-minute conversation holds far more than one post. Most teams publish the episode, cut one clip, and move on.

In Issue 4, a podcast editor explained why shows stall around episode 10. The ones that keep going treat every recording as a library of clips, not a single upload. See how Leo AI edits your video into scheduled social posts.

Written by Lera Leonteva. Prices are US street prices checked in September 2026 at B&H, MPB, Sweetwater, Focusrite, Sigma, Amaran, Neewer and Ulanzi; camera prices have moved with tariffs this year and used prices vary by condition. Recording limits from Sony’s help guides and recordinglimits.com. Room prices are New York rates from WeWork, Regus, Deskpass, LiquidSpace and Peerspace.

Frequently Asked Questions

What teams ask before they buy their first light.

What is the cheapest camera setup for a professional-looking video podcast?

A used Sony a6000 with a TTArtisan or 7Artisans 35mm f/1.4 lens, about $490 per camera in September 2026. The camera matters less than the light: a gridded key light and a dynamic mic close to the mouth do more for the look than a more expensive body.

Why does lighting matter more than the camera?

The expensive look is a face lit softly from one side while the background falls into shadow. That comes from a key light in a softbox with a honeycomb grid, which keeps the light on the speaker and off the wall. An expensive camera in bad light still looks cheap.

What microphone should I use for a video podcast?

A dynamic microphone a hand's width from the mouth, because it picks up the voice and very little of the room. The Fifine K688 and Samson Q2U connect by USB or XLR, so a solo host can plug straight into a laptop. The Rode PodMic is XLR only and needs an interface such as a Focusrite Scarlett Solo, or a Vocaster Two for two mics.

Do budget cameras have recording limits?

Yes. The Sony a6000 stops every 29 minutes and can overheat on long takes, so restart at natural breaks and power it with a dummy battery. The Sony ZV-E10 records HD for up to about an hour with its overheat setting on High. The Canon EOS R50 records up to 59:59 per clip.

Can I book a coworking meeting room without a membership?

Yes. In New York, WeWork On Demand rents rooms by the hour from $8 a seat, Regus meeting rooms start at $35 an hour in Manhattan, Deskpass from about $30, LiquidSpace from about $50, and Peerspace lists rooms and podcast studios at $35 to $100 an hour, usually with a 2-hour minimum.

How far should I sit from the back wall?

Six to ten feet. Distance is what blurs the background. Sitting against the wall flattens the shot even with a good lens.

Lera Leonteva
Lera Leonteva
CEO & Founder, Leo AI

Lera Leonteva is the founder and CEO of Leo AI, where she works on turning long-form footage into short-form video that earns attention on its own merits. She writes Video Scaling Strategies, taking one piece of content that worked and one that did not, and using what the market is doing to explain the gap.

Have a video?

Leo AI edits it into scheduled social posts, guided by what your audience actually watches.

See this working on your own channels

Leo turns your existing footage into posts and shows what each one earned. Book a demo or see the channels Leo posts to.