The distribution problem for a factory of 220 apps is different from the distribution problem for one app. A single app optimizes a single message for a single audience. A factory has 220 audiences, and each one needs its own message — but if the messages can share infrastructure, the economics work in a way they don't for a single product.

Short-form video is the channel where this works. Here's the thesis.

The Channel Comparison

Look at how different distribution channels scale with catalog size:

The last one is the unlock. Every other channel costs something per vertical. Video costs once per vertical (making the video) and then the algorithm does the rest.

Why Short-Form Specifically

Three properties of TikTok / Reels / Shorts that matter:

  1. Algorithmic targeting. You don't need to pay to target beekeepers — the algorithm figures out that a video about hive logs goes to beekeepers. The targeting is free.
  2. Low production cost. A 45-second video about honeybees shot on a phone, voiced over, with clean text overlays, is a 30-minute project. 220 verticals at 30 minutes each is 110 hours. Stretched over 6 months, that's 4 hours/week of video production.
  3. Niche discovery. Users on short-form platforms are actively browsing niches. Someone who watches three cheesemaking videos is getting served my cheesemaking video at high probability. The feed mechanism finds my audience for me.

The Production Pipeline

The plan I'm working toward looks like this:

  1. Script. 45-second script per vertical, generated from the spec. "Here's the problem {hero_em} have. Here's what our app does. Here's where to get it." Claude can first-draft these from the spec; I edit.
  2. Voice. Record the voiceover on my phone. No talking-head video; just b-roll + text overlays + voice.
  3. B-roll. Stock b-roll of the relevant context (cheesemaking process, beekeeping inspection, tattoo studio, whatever matches the vertical). Licensed footage or self-shot.
  4. Text overlays. Captions and key points rendered in CapCut or similar. Branded, but not heavy-handed.
  5. Post. TikTok + Reels + YouTube Shorts simultaneously. Same video, cross-posted.

Batch production: 5-10 videos in a single Saturday. 220 videos in 6 months of Saturdays.

The Math I'm Hoping For

Conservative assumptions:

660 posts × 2,000 views × 2% × 5% × 5% = 66 paying customers. Across the catalog. From one pass.

Optimistic assumptions (some videos go viral, average view count 15,000):

660 × 15,000 × 2% × 5% × 5% = 495 paying customers.

Somewhere in between is plausible. Let's call it 100-300 paying customers generated by 6 months of video production. At average $45/month per customer, that's $4,500-13,500/month of additional MRR from one content initiative.

The Compounding Piece

Short-form video posts don't decay the way cold emails do. A TikTok from six months ago is still getting views today. The library compounds. Post 220 videos; a year later, you have 220 pieces of evergreen content still pulling in new leads at essentially zero ongoing cost.

Content marketing via blog posts has the same compounding property but at much longer time horizons. Short-form video gets the compounding benefit on a 6-month cycle instead of a 3-year one.

The Risks

Three ways this doesn't work:

Algorithm change. TikTok/Meta/YouTube could shift their algorithms toward longer-form, celebrity content, or paid promotion. If that happens, the "free targeting" property goes away.

Production burnout. Making 220 videos, even at 30 minutes each, is real work. If the first 50 don't show traction, I'll struggle to make the next 170. I need to validate early and either pivot or commit.

Wrong-audience views. 2,000 views from the wrong audience is worse than 200 views from the right audience. The niche-targeting only works if the platform identifies the right niche for each video. Early videos might not get algorithmic traction if the signals aren't clear.

What I'd Skip

Not doing: in-video personalities, heavy editing, stock music, talking-head. These add production cost without proportional conversion lift for B2B SaaS content specifically. The audience for a cheesemaking batch tracker wants to see the product, not me.

Not doing: TikTok Shop, influencer partnerships, paid boosts. The thesis is organic reach. Paid kills the unit economics of a 220-vertical catalog.

The cheapest channel is the one where your structural advantage lines up with the platform mechanics. For Dangercorn, that's short-form video over algorithmic discovery. For most SaaS companies, it isn't.

Why Not Just Buy Ads

The obvious alternative is paid advertising. Google Ads on niche keywords. Facebook ads to lookalike audiences of confirmed customers. Why not just spend money?

Because for niche B2B SaaS at our pricing, paid acquisition rarely pencils out. CPC for keywords like "cheesemaking software" is $2-4. Conversion from click to landing-page-view is 100%, but conversion from landing to signup is maybe 3%, and from signup to paid is maybe 5%. That's $2-4 / (3% × 5%) = $1,300-$2,700 customer acquisition cost. At $39/month average, payback is 33-69 months. Brutal.

Organic content (video, blog, SEO) has higher upfront cost but near-zero marginal cost per acquired customer once the content exists. The unit economics are dramatically better at our scale. Paid would work if we had higher LTV; we don't, so we don't.

Related

Why cold email isn't scaling. The 220-app thesis. Why niches are the moat. The economics that benefit from more paying customers.