What a view costs and how to handle with it

Advertisers have known their number for twenty years. It's called CPM — what you pay for a thousand impressions. Organic people have no equivalent, mostly because the invoice never arrives. Nobody bills you for the Tuesday night you spent editing.

Let's send the invoice.

It's not complicated. That's the problem — it's simple enough that everyone could have run it and nobody has.

Three inputs. Two of them you already know. The third is the one you've been avoiding: what your hour is worth.

If you're a founder who could be billing $75 an hour, or shipping features worth more than that, your hour costs $75 whether or not money changes hands. Every hour of it. This is the same arithmetic that makes "I'll just build it myself" sound cheaper than it is, and we'll come back to that at the end, because it turns out to be the whole point.

Median views from our own data. Paid CPMs from published 2026 benchmarks. Three production scenarios, from filmed it on my phone in the kitchen to actually made something.

At two hours and $100/hour — $200 a post: the Instagram number is $1,613. The ratio is 210x. You can see where this goes.That distinction is the whole game right now: a single-purpose bridge versus one that carries your entire social and ad operation across a single connection.

Flip it around and it gets worse. To make a 45-minute post pencil out at TikTok's paid CPM, that post needs 6,141 views. Median is 401. You're not asking for a good post. You're asking for a 15x-above-median post, every time, forever.

No. And if you stop reading here you'll draw exactly the wrong conclusion, which is why most people never run this math in the first place — it's dangerous in isolation.

CPM is a rented metric. It measures one thing: the price of being seen once, by someone who did not ask to see you, and who you will have to pay for again tomorrow. Stop the spend and the impressions stop that afternoon.

Three things the organic number doesn't capture, and they're not soft:

The view isn't the deliverable. A paid impression buys attention. An organic post buys attention plus a follower who sees the next one for free, plus a post that keeps surfacing, plus a piece of evidence that you know what you're talking about. The ad expires. The post is an asset with a residual.

You can't buy the thing that makes organic work. Nobody has ever trusted a company more because of a display ad. The reason organic distribution is expensive is the same reason it works: it costs you something visible, and the audience can tell.

The distribution is a lottery, and lotteries don't have medians that matter. We showed this last time — the top 1% of posts capture 78–90% of all likes. Your median post costs $454 per thousand views. The one that breaks out costs a rounding error per thousand. You're not buying views. You're buying tickets, and the expected value lives entirely in the tail.

Which means the right response to this table isn't to abandon the channel. It's to attack the two variables in the formula.Put together, a founder running a small team can now say something like: "check my TikTok performance from the last 28 days, pause anything with cost-per-result climbing, and boost my top organic video as a Spark Ad with a $100 budget, paused for my approval" - and get exactly that, instead of spending an hour clicking through a dashboard to do it by hand.

You can't do much about your hourly rate. You can do a lot about the other two numbers.

5.9x cheaper per view for zero additional production. Nothing else in our entire dataset moves a number that far for free.

Stack both — video format, four destinations — and the median post goes from 59x worse than ads to something in the neighborhood of 10x. Still not parity. But now you're paying a premium for an asset that compounds instead of setting money on fire for one that doesn't.

Read the table again and notice what almost happened to you.

You saw a cost-per-unit number, compared it to a cheaper cost-per-unit number, and nearly concluded the expensive one was wrong. Everyone does this. It's the most natural mistake in business, and it is the same mistake — structurally identical — that is currently wrecking how you price your product.

Because cost-based thinking is a trap on both ends of the business.

On distribution, it tells you to stop posting — because the CPM is bad — when the CPM was never measuring what you were buying.

On pricing, it tells you to charge less — because your build cost collapsed — when your build cost was never what the customer was paying for.

Here's how that plays out. It's 2026. You built the thing in a weekend with AI. Your marginal cost is a $40 infrastructure bill. So you price it at $29 a month, because charging more for something that took you a weekend feels like fraud.

That instinct is the same arithmetic error, wearing a different hat.

Your customer's alternative is not your competitor's pricing page. It's their own hands. So run their version of the formula: hours × their rate. A developer who "just builds it themselves" is looking at the same phantom-free-labor illusion as the founder who thinks organic posting costs nothing. Two weeks of build time at $100/hour is $8,000 — before anything breaks. Your $29 price tag isn't competing with $0. It's competing with $8,000 that they've decided not to count.

The build is a one-time cost. The maintenance is a CPM — it recurs, forever, at a rate you don't control. Rate limits change. OAuth flows change. An endpoint gets deprecated on a Thursday with 30 days' notice. What you're selling isn't the feature; it's the permanent absence of that Thursday. Price the recurring thing recurringly.

The reason organic distribution survives a 59x cost disadvantage is that it accrues. Same test for your product: if what you sell gets more valuable to the customer the longer they use it — accumulated data, accumulated integrations, accumulated switching cost — then flat pricing is you handing that compounding away for free. Find the unit that grows with their success and price on that.

Both count the cash and ignore the hours. Both are wrong by roughly an order of magnitude. And both get corrected by the same move: put a number on your time, then decide.

Most people never run the number, because the number is uncomfortable and nobody sends the invoice. Run it anyway. It's the only way to find out which of your free things are actually expensive, and which of your expensive things were the bargain all along.If you're already testing this, hit reply and tell me which platform you're wiring up next: TikTok, Meta, or something less obvious.

Organic view data — Median views per post from Zernio production analytics, the same 36.06M-post dataset published on July 29: Instagram all posts (124 median views), Instagram Reels (160), Instagram non-Reels (48), TikTok (401). Medians, not averages, for the reasons covered in that issue.

Paid CPM benchmarks — Published 2026 figures, using conservative values: Instagram $7.68 (WebFX 2026 Meta benchmarks, range $6.25–$7.68); TikTok $9.16 in-feed (2026 pricing data, planning range $4–$13). We deliberately used the high end of the Instagram range, because it's the number least favorable to the argument being made.

Production cost — Our assumption, not measured. Three scenarios spanning 15 minutes to 2 hours, at $50–$100/hour. Substitute your own; the formula is above.

Crosspost bundle — Sum of published medians for Instagram all-posts, Instagram Reels, TikTok and Instagram non-Reels as a four-destination stand-in (~733 views). Illustrative, not a measured four-platform cohort. Real crossposting results depend on whether the asset fits each destination — a 9:16 TikTok cut does not perform as a LinkedIn post, and pretending otherwise is how people conclude crossposting doesn't work.

Not included — LinkedIn, X and Threads. We publish likes data for those platforms but not comparable per-post view medians, so there is no honest cost-per-view to report. For reference, LinkedIn's paid CPM runs $28–$33 globally, which means the organic-vs-paid gap there is structurally different from the two platforms above.

Known limitations — Ad "impressions" and organic "views" are counted differently by every platform and are not strictly equivalent units; treat cross-platform ratios as order-of-magnitude, not decimal-precise. Our user base skews to creators, SMBs and developers, which is exactly why these medians are lower — and more honest — than industry benchmark reports built on brand accounts.