Introduce a New Product Version Without Making the Current One Look Unusable

The first cut of a launch film rarely fails because it lies. It fails because it compresses. A hinge moves two centimetres, and the edit language becomes "the problem with the old design." A sensor gets faster, and the voiceover says "finally, a light you can trust." Nothing in those lines is a discrete falsehood. Together they assert a history the product never had, and the people who bought that history are watching.

Separating what changed from what the change is worth is most of the work. The rest is refusing the dramatic shortcut that makes the old version look broken, because that shortcut is usually cheaper to produce than the explanation it replaces.

Establish the version facts before choosing the drama

Before you decide whether this is an improvement story or a replacement story, write down the comparison in plain terms. Not the marketing claim — the actual difference between what shipped and what now ships.

A working inventory, for the fictional folding task light I'll use throughout:

Read that table and two rows have moved. The functional change is the hinge angle — from flat or open to roughly 60° above the bench, via the raised stop. The colour range changes too: still four options, but two of them new, so the v1 palette doesn't carry over whole. Clamp, cell and driver are continuity. The temptation, once you have a table like this, is to look for drama in the empty cells — to imply the v1 hinge was somehow deficient because it lacked a position nobody asked for. Resist that. An absent feature is not a past failure.

Two discipline questions belong here, and they belong to other people. Does the v2 hinge exist in every unit and every market at launch, or is it a configuration? Is the raised stop compatible with existing clamps and accessories, or does it change where the head can reach? Those are availability and compatibility questions for the product owner, not for the treatment. Write the film around the answer you're given; don't write the answer into the film.

Show the improvement at the scale it actually is

Here's the same change told two ways. Both are for the same v2 light. Watch what each one claims about the world.

Bounded version, first pass:

A workbench at night. A hand brings the v1 light in, opens it, and the beam falls across a scattered row of small components. The hand moves one piece to the side, out of the glare. Cut to the v2: the head tips up onto the raised stop, and the same row sits in even light with the components' faces visible. Voiceover: "The new stop holds the head at about 60 degrees, so the light reaches across the bench instead of pooling on it."

Replacement version, first pass:

Same bench. A hand brings in the v1 light, opens it, and the beam pools uselessly on a heap of parts. The hand scrabbles at them, can't tell one from another. Voiceover: "Stop fighting your light." The v1 is set aside, almost dropped. The v2 arrives, and the bench becomes legible.

The replacement version is doing something the bounded version isn't. It asserts that the v1 cannot do ordinary close work — that it "pools uselessly." But the v1's two positions worked for the tasks it was made for. In the bounded scene, the v1 is doing fine until the job changes: the components need to be seen in a row, not lit individually. That's the honest shape of the improvement. The new position helps when the work is spread out; it doesn't retroactively make near work impossible.

Notice the small machinery. The bounded scene moves a piece "out of the glare" — a real, minor friction, not a failure. The replacement scene has the hand "scrabbling," can't distinguish parts, nearly drops the light. Every one of those gestures is doing argumentative work: they're manufacturing incapacity. The v1 was never represented as lighting a scattered row evenly. Making it fail at a task it never claimed is the cleanest way to slide an unsupported claim past a viewer, and the easiest way for a v1 owner to catch you.

What the replacement story would actually require

A replacement story isn't automatically wrong. There's a real genre where the new version does supersede the old: a changed requirement makes the previous tool genuinely unfit. If v2 had a new mounting standard, or a different power architecture, or a regulatory constraint that made v1 non-compliant, the dramatic discard could be earned.

But look at what that demands. A "changed requirement" replacement story needs facts that establish the change is real and current: what requirement changed, when, for whom, and under what authority. It needs the old product's limits stated specifically — not "can't handle the work" but "cannot meet the new mounting standard in the regions where it applies." It needs a defined population: all v1 owners, or those in a category, or those doing a particular job. And it needs the v2 to actually satisfy the new requirement everywhere the story implies, which loops straight back to the availability and compatibility questions from the earlier section.

None of that is inventable in a treatment. A provisional script can pose the replacement structure; it cannot supply the changed requirement. If the product team gives you a verified change — a standard, a spec, a legal date — the discard can be honest. Absent that, "stop fighting your light" is a mood, not a claim, and the film will be working against its own evidence.

Compare the routes and what each one costs you

Set the two concepts side by side as production choices, because that's what they are.

The bounded improvement story is precise and cheap to defend. Every claim maps to a verified difference; the v1 appears doing real work; existing owners see their tool respected. Its cost is explanatory. You have to make the use case legible — the scattered row, the need to see faces not just spots — and that means the film spends running time establishing a problem-in-use rather than a problem-in-the-product. It's a slower open. Some clients will feel it lacks "impact," which usually means it lacks a villain.

The replacement story is dramatic and expensive to defend. It carries the burden of the changed-requirement evidence, plus a population definition, plus v2 availability. If any of that is soft, the film makes a promise the product can't keep, and the audience most likely to notice is the one that already owns v1.

You can't blend them. A cautious voiceover over images of universal obsolescence doesn't hedge — the images win. If the voiceover says "an added 60° position for spread-out work" while the picture shows the old light flickering out, the picture is the claim. Pick the route the facts can carry, and don't let the edit reverse it.

Preserve continuity without writing the customer's mind

What remains familiar is part of the offer, and you can say so without narrating anyone's feelings. The clamp is unchanged. The cell and driver are unchanged. An existing owner's accessories still fit. That's continuity you can show — same clamp action, same familiarity — and it's a factual statement, not a reaction.

What you can't do is assign the past buyer an emotion. No "you'll wish you'd waited." No "finally, the light you deserve," which quietly tells everyone who bought v1 that they were foolish. No "upgrade and stop struggling." These aren't just unkind; they're unsupported. You don't know that v1 owners struggled, and you have no evidence for regret, gratitude, or relief. Regret is a claim about a person, and it needs evidence the way any other claim does.

So keep the two registers separate. Hypothetical use situations — "imagine a bench where the parts need to sit in a row" — are fine and clearly marked as such. Audience research is a different thing entirely, and if you don't have it, don't dress a hypothesis in its clothes. And leave the unresolved conditions visible: if the raised stop's reach is still being confirmed, don't shoot a scene that resolves it. A held question in the treatment is cheaper than a reshoot after the spec changes.

The last line should name what's better and what isn't

End with the account the facts support, in the order the work will use it: what's better, for which use, and what hasn't changed. For the task light, that closing line might read: "The head holds at roughly 60 degrees, so spread-out work sits in even light; the clamp, cell, driver and accessory fit are unchanged." Three clauses, no villain.

Then choose the story those clauses can carry. The bounded improvement is usually the one the facts will hold, and its slower open is the price of a claim you can defend in front of the people who bought the last version. A replacement story is available whenever the change is real and verified — and when it isn't, the discard sequence isn't a bolder film. It's a film that spends its credibility in the first ten seconds and never earns it back.

Frequently asked questions

How can we tell whether a launch film is presenting an improvement or unfairly making the old version look broken?

Compare every claim with the verified version facts. An absent feature is not a past failure. A bounded improvement shows the old version doing real work until the job changes, while a replacement story asserts the old version cannot do tasks it was made for. If the old product is made to fail at a task it never claimed, that is an unsupported claim.

What does a replacement story require before it can honestly discard the old version?

It needs a real, verified changed requirement: what changed, when, for whom, and under what authority. It also needs the old product's limits stated specifically, a defined population, and evidence that the new version satisfies the requirement everywhere the story implies. Without a verified change, the discard sequence is a mood rather than a claim.

What are the tradeoffs between the bounded improvement and replacement routes?

The bounded improvement is precise and cheap to defend: every claim maps to a verified difference, and the old product appears doing real work. Its cost is explanatory running time, because the film must establish a problem-in-use. The replacement route is dramatic but expensive to defend: it carries the burden of changed-requirement evidence, population definition, and new-version availability. If any of that is soft, the film makes a promise the product cannot keep.

Can we hedge by pairing a cautious voiceover with images of obsolescence?

No. The routes cannot be blended because the images win. If the voiceover says an added position serves spread-out work while the picture shows the old light flickering out, the picture is the claim. Pick the route the facts can carry, and do not let the edit reverse it.

How should the film handle existing customers and unchanged features?

Show continuity factually: unchanged clamp, cell, driver, and accessory fit are part of the offer. Do not assign the past buyer an emotion such as regret, gratitude, or relief; those are unsupported claims about a person. Hypothetical use situations are fine when clearly marked, but audience research is a different thing. Leave unresolved conditions visible rather than shooting a scene that resolves them.

Further reading

• The Required Claim Needs More Reading Time Than the Scene Allows — Read this when a required condition has to fit inside the launch film and viewers need enough time to understand it before the reveal.
• Make the Commercial’s Hero Product Look Like the Product People Will Receive — Consult this when preparing hero frames or renders for the launch, so the product shown matches the version buyers will actually receive.
• Research a Product’s Real World Before Looking at Other Ads — Use this before writing the bounded demonstration, to ground the use case in observed behavior rather than a generic scene or borrowed campaign look.