Repeatable format · Weekly, and only in weeks you actually shipped

    Run a “What We Fixed This Week” Series

    Turn shipping into a format — and turn the thing you're still bad at into the reason people believe the rest.

    Cadence

    Weekly, and only in weeks you actually shipped

    Runtime

    20-40 seconds

    Beats

    5

    Episode bank

    10 angles

    The spine: The count card, the broken-then-fixed cut and the “still broken” slot are fixed. The fixes are the variable.

    A changelog is the least glamorous thing a company publishes and one of the most trusted, because it is the rare piece of marketing that can be checked. Turning it into a weekly video format keeps that property and adds the one thing a written changelog lacks: you can see the difference.

    The mechanic is show-don't-tell, applied literally. Two seconds of the broken state followed by two seconds of the fixed state communicates more than a sentence describing the improvement, and it takes less time. A described fix is a claim; a shown fix is evidence.

    The beat that carries the entire format is the last one — naming something that is still broken. Every brand publishing improvements sounds identical; the one that also publishes its open problems is the one people believe, and it costs nothing except the discomfort of saying it.

    Best for: Products that change — software, services, menus, processes

    The episode skeleton

    Same beats every week. The timings are a starting shape, not a rule — what matters is the order and that no beat gets dropped.

    1. 1

      The count

      0-3s

      “Three fixes this week.” A number is a promise the viewer can hold and it sets the length expectation, which is why this format retains better than an unnumbered update.

    2. 2

      Fix one, shown

      3-12s

      Two seconds of the broken behaviour, then the fixed one, in the same framing. Show the annoyance before the solution — a fix to a problem the viewer hasn't felt reads as nothing.

    3. 3

      Fixes two and three

      12-28s

      The identical shape, each one shorter than the last. Descending length keeps the pace moving and puts your strongest fix where attention is highest.

    4. 4

      What's still broken

      28-35s

      Name one open problem, without a promised date unless you mean it. This is the entire credibility of the format and it is the beat everyone wants to cut.

    5. 5

      Next week

      35-40s

      The same closing card every week. It makes the series feel like an obligation you've taken on, which is precisely the impression that builds trust.

    Decide once, then never again

    This split is what makes a format cheaper every week instead of the same price every week.

    Fixed forever

    • The count card and its position at the top
    • The broken-then-fixed cut, in matched framing
    • The “still broken” slot at the end, every episode
    • The closing card and the weekly slot

    Changes each episode

    • The fixes themselves, and the open problem you name

    10 episodes to start with

    Adapt each one to your own category. At the stated cadence this is roughly two months of the series before you have to think of anything new.

    1. 01A bug a customer reported, credited with their permission
    2. 02The confusing wording you rewrote
    3. 03A step you removed from onboarding entirely
    4. 04The thing that was slow and now isn't
    5. 05A change you shipped, then reverted, and why
    6. 06The support macro you retired because the problem is gone
    7. 07A default setting you changed, and what it was costing people
    8. 08Something you fixed that nobody had asked for
    9. 09The fix that took three attempts to get right
    10. 10The oldest open issue you finally closed

    How to build it in Versely

    Every format on this hub is honest about how close the published library gets. This one has no workflow that runs in this shape — so the instruction is to clone the closest one and swap its asset map.

    Clone the closestNoseFresh — Nasal Strips Ad15 scenes · 1 asset

    Nothing published is a changelog. This is the closest structure — a problem-then-solution arc with explicit before-and-after cutaways where every earlier scene gets a callback once the fix lands. Clone it and swap the asset map, then repeat the problem-to-fix beat once per item instead of stretching a single problem across the whole runtime.

    Caption preset

    Classic Top

    A plain white caption at the top of frame stays out of the way of the interface, product or process being shown, which matters in the one format where the footage carries the information and the caption is only labelling it.

    Once the first episode exists, the series itself is a schedule rather than a task: Run a recurring video series. Set it once. It writes a new version every time it runs.

    Keeping it honest

    Only claim fixes that actually shipped to customers, and don't credit a customer by name without asking first. If you name a date for an open problem, treat it as a commitment — a missed date in a format built on candour costs more than never having given one.

    When to kill it

    A format that has stopped working and keeps publishing is worse than no format at all. These are the signals to stop.

    • Nothing shipped. Skip the week — padding a changelog with cosmetic changes is how the format stops meaning anything.
    • The “still broken” slot goes empty two weeks running. Either you've stopped being honest or you've stopped having ambitions.
    • Fixes get described instead of shown. Once it's a person talking about improvements, it's an update video and it will be watched like one.

    Frequently asked questions

    Isn't publishing our bugs bad marketing?+

    Publishing that you fix bugs is good marketing; the bugs were already being experienced by the people who had them. What changes is whether those customers see a company that noticed. The risk is in the pattern, not the disclosure — a changelog that shows the same class of problem every week is telling a story you should fix before you publish it.

    What if we ship too slowly for a weekly cadence?+

    Move to fortnightly or monthly rather than padding. The count in the first beat is what makes the format work, and a week with one real fix and honest framing beats a week with three cosmetic ones.

    Who should present it?+

    Whoever did the work, if they're willing. This format is unusually resistant to polish — a slightly awkward person showing a real fix outperforms a produced presenter describing it, because the point being made is that a person did this.

    How do I show a fix that isn't visual?+

    Show its consequence rather than the change. A speed fix is a stopwatch on the old and new versions, a wording fix is the two sentences side by side, a process fix is the number of steps before and after. If there is genuinely nothing to show, it belongs in a written changelog instead.

    Other formats worth running

    Missing something you’d need to shoot this?

    No camera, no studio, no presenter, no script — each gap has a route that still gets the episode made. The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.