Pitch a Campaign That Teaches a Product One Step at a Time

The instinct behind a "how-to" campaign is usually sound and the execution usually isn't. Five films, five features, numbered. It looks like a curriculum because of the numbers. Then you ask which film someone sees first, and the honest answer is whichever one the algorithm serves them at 11 p.m., which means every film has to stand alone anyway, which means the numbering was decoration.

So the useful move at pitch stage is to stop counting films and start mapping what each one assumes. A teaching campaign isn't a list of lessons. It's a route through prerequisites, and a route only exists if something preserves its order — the films, the placement, or the viewer's own reasons for seeking them out in sequence.

That distinction is the whole pitch. Everything below is one way to make it visible on paper, using an invented example. The product, its interface, and every delivery assumption here are fictional and hypothetical; they are a demonstration of reasoning, not a tested campaign.

Give each film a bounded action, not a feature

Start with what the viewer can do after the film that they couldn't do before. Not "understands Collections." Not "is aware of the Reorder tool." An action you could watch someone perform.

Vague units are where campaigns quietly become feature tours. If Film 3's stated job is "introduces sharing," you have licensed yourself to show the share sheet, mention three destinations, and end on a logo. Nothing in that film tells you what has to be true before it can land.

A bounded action does. Here's the fictional sequence:

Film 1 — Place images in a collection. Viewer opens the app, selects three or more photos from the library, and confirms them into a named collection. Success state: the collection exists and is visible.

Film 2 — Reorder the collection for a story. Viewer opens an existing collection and drags images into a deliberate sequence. Success state: the order is saved and differs from import order.

Film 3 — Share the collection. Viewer opens an existing, ordered collection and shares it through one simulated route. Success state: a share confirmation appears, and the recipient-facing view reflects the order.

Three films, not five, because I couldn't defend a fourth or fifth action that wasn't a feature mention in disguise. That's the first thing a pitch should be able to say out loud: this number, and no more, because these are the actions and the rest were ornaments.

Now the important part. Each of those success states names something that has to exist first.

Map the prerequisites before you map the running order

Take Film 3 and ask, literally, what must already be true.

The viewer must have a collection. That collection must be ordered, or the share is just a dump. And the viewer must know that a collection can be shared — not where the button is, but that the feature applies to this object at all. That last one is the prerequisite most campaigns skip, and it's usually the one doing the actual work.

Film 2's prerequisites: an existing collection with images in it. The viewer also has to believe that order is theirs to set, which is not obvious in an app that might sort by date by default.

Film 1's prerequisites: images in the library, and enough working knowledge of the app to find the library. That's the floor.

So the dependency graph, which is the artifact I'd actually put in the deck:

Two things fall out of writing it down.

First, the escalation is real, not cosmetic. Film 3 cannot be authored as "here's the share button" without silently requiring the viewer to have done Films 1 and 2. If you show a share sheet over a populated, ordered collection without acknowledging where that collection came from, you've made a film whose first frame assumes the last two.

Second, the "nice to know" column is where campaigns smuggle in prerequisites that aren't. Sync settings aren't needed to place an image. Cover-image behavior isn't needed to reorder. Cut them, or they'll crowd the film and make the actual action harder to see.

The transition is the argument

Between Film 1 and Film 2 there's a hinge that decides whether this campaign is coherent. Placing images into a collection feels finished. The collection exists, it's named, it's there. Nothing on screen says "you are one step into a three-step job."

Film 2 has to reopen that closed feeling without re-teaching Film 1. Hypothetically, the opening might look like this:

Open your collection. It's in import order — the first photo you added is first. But a collection you share isn't a folder. It's a sequence someone reads. So this film is about the order.

That's four sentences doing prerequisite work: it assumes the collection exists (Film 1 delivered it), confirms it's the same object, names the new capability, and frames why it matters. A film that instead re-explains collections has wasted its runtime on material the campaign already handled. A film that skips the reorientation has assumed the viewer watched Film 1 and remembered the exact object it produced — neither of which the campaign controls. What it controls is the artifact: that Film 1 exists and teaches that object. It cannot guarantee the viewer watched it, or that they remember it, which is precisely why the reorientation line has to carry the object itself rather than gesture back at it.

This hinge is what a pitch should be judged on. Anyone can list three features. The defensible claim is: this is where the sequence earns its order, and here is the line that carries it.

Then check whether your delivery can actually hold an order

Everything above is authored order. It says nothing about viewing order, and the gap between those two is where most teaching campaigns quietly fail.

If the placement is three paid social films served independently to overlapping audiences, there is no sequence. There is a set. The films will be encountered in whatever order, at whatever times, by people who saw zero, one, or all of them, in any configuration. If the placement is a course module, an in-product walkthrough, or a landing page where the films sit in a visible order and the next one is the obvious click, you have something closer to a route — but "closer" is a claim about the plan, not about learning.

I can't tell you which you have. That depends on your actual environment, your placement specs, and what the platform currently does with delivery order — all of which need checking against current documentation, because it changes and because a plausible-sounding guess is not a plan. What I can tell you is what to write in the pitch once you know: a single sentence naming the delivery assumption and what it does and doesn't guarantee.

Something like: Ordered as a course in the product's help center; viewers arriving from the standalone social cut of Film 3 must be able to complete the action without Films 1 and 2. That sentence has a cost attached, and naming the cost is what makes the rest of the pitch credible.

The fork: sequential or self-contained

Once you know what the delivery can preserve, you get a real choice, and it has different answers for different films.

Where order holds: later films build. Film 2 opens with four lines of reorientation, not a re-teach. Film 3 assumes both prior actions and spends its time on the share itself. The campaign gets shorter and more confident, and the viewer gets a genuine progression rather than three parallel introductions.

Where order doesn't hold: each film has to be a legitimate entry point. That means Film 3's version of the story includes the collection. Concretely, it opens on a populated, ordered collection and says how it got that way in one clause — "photos you've placed and sequenced" — so the viewer knows the object exists, knows it's theirs to make, and knows this film isn't going to teach that part. Then it either shows the share or points at where the earlier steps live.

That's the honest trade. Self-contained units cost repetition and runtime, and they still don't produce learning — they produce awareness of a sequence that exists somewhere else. A viewer who sees the standalone Film 3 knows sharing exists. They do not have a collection.

A third route is often stronger than either: keep the films self-contained but make one of them, or an accompanying page, the structured resource — a single place where the three actions appear in order with their prerequisites visible. Then the standalone films have somewhere to send people, and the ordered sequence has somewhere to live, which means an uncontrolled entry point can hand the viewer a route rather than leaving them to assemble one. This is a delivery decision, not a film decision, and it belongs in the same conversation.

Shorten the film without deleting the operation

The last trap is trimming. A teaching film gets cut for time and what disappears first is the step that felt obvious to whoever was editing — usually a confirmation, a save, the moment where a drag actually lands.

Check each cut against the success state. If Film 2's stated action is "the order is saved and differs from import order," then a version that shows dragging but cuts before the save is no longer teaching that action. It's showing two images swap places. The film still looks like a how-to. It just doesn't contain the step the viewer will fumble.

Keep the steps accurate to the actual product. If the interface changes, the film is wrong, and the pitch should say who owns keeping it current. And keep the scope honest: this method produces a clear route through a few low-stakes actions. It is not safety training, it doesn't certify anyone, and a viewer who watches three clear films is not thereby competent. Say that in the deck rather than letting a clean sequence imply a guarantee it can't make.

What to put on the page

Two artifacts and a sentence.

The map: action, success state, genuine prerequisites, and the nice-to-know column you're deliberately not filming. Three rows here, but the format holds at seven.

The route: where the films sit, in what order they're visible, which entry points are uncontrolled, and — for each film — whether it's sequential or self-contained. If a film is self-contained, write the clause that restores its starting context. If it's sequential, write the line that reorients without re-teaching. Those lines are the campaign.

And the sentence about what order your delivery does and doesn't preserve, with the assumption marked as an assumption.

If you can't establish an order, don't paper over it by numbering the films. Change the entry conditions instead. Either make each film stand on its own, or build the one place where the sequence actually holds — and then send people there. Numbering a set of independent films is the one move that teaches nobody anything, and it's the easiest one to pitch.

Frequently asked questions

Why does the example stop at three films?

Because a fourth or fifth action could not be defended as anything other than a feature mention in disguise. Each film is given a bounded action with a watchable success state — place images in a collection, reorder it, share it — rather than a vague unit like introducing sharing, which licenses showing a share sheet, naming a few destinations, and ending on a logo.

What distinguishes a genuine prerequisite from nice-to-know material?

The table separates them. Placing an image requires images in the library and the ability to navigate it; cloud sync and storage limits are not required. Reordering requires an existing collection with at least two images and the belief that order is manually settable; auto-sort rules and cover-image behaviour are not. The nice-to-know column is where campaigns smuggle in non-prerequisites that crowd the film and obscure the actual action.

Will numbering the films create a sequence?

No. A route exists only if something preserves its order — the films, the placement, or the viewer's own reasons for seeking them out in sequence. Three paid social films served independently to overlapping audiences are a set, encountered in whatever order, at whatever times, by people who saw zero, one or all of them.

What does it cost to make every film self-contained?

Repetition and runtime, and it still does not produce learning — it produces awareness of a sequence that exists somewhere else. A viewer who sees only the standalone third film knows sharing exists; they do not have a collection. The reorientation clause has to carry the object itself, for example naming it as photos you have placed and sequenced, rather than gesturing back at earlier material.

Does a clear three-film sequence make viewers competent?

No. The method produces a clear route through a few low-stakes actions. It is not safety training, it certifies no one, and a viewer who watches three clear films is not thereby competent. That should be said in the deck rather than left for a clean sequence to imply a guarantee it cannot make.

Further reading

• Tell a Commercial Through Interface Actions, Not a Tour of Screens — Read while drafting the films themselves: it shows how to write interface steps as actions that change someone's situation rather than a numbered tour of screens.
• Pitch a Series Viewers Can Enter at Almost Any Episode — When a film must work as a first encounter, borrow its method of auditing an ordinary installment for the facts a stranger needs to follow what happens.