Before/After Explainer Structure That Still Teaches

Before/After Explainer Structure That Still Teaches

Structure transformation teach Reels as before → mechanism → after so viewers learn the change—not just admire a vanity glow-up.

Empty before/after is everywhere in feeds: messy desk → tidy desk, red chart → green chart, with no transferable step. Educators and product marketers shipping vertical teach content need the middle act. Pair this beat map with B-roll vs demo UI choices per beat so each act gets the right visual type.

Why vanity before/after fails teaching

Vanity frames optimize for scroll-stop, not skill transfer. They borrow the grammar of transformation ads without paying the cost of instruction.

If the after state could belong to any tool in your category, the Reel taught brand mood, not process. That is fine for awareness ads. It fails for teach content that should earn saves and replays.

A second failure mode: the "before" is cartoon misery and the "after" is utopia. Extreme contrast feels false in B2B and regulated niches. Prefer a realistic broken state the viewer already recognizes from their week.

The three-act beat map

Lock these acts on the board before any clip renders:

1. Before (stakes + specific failure) — one concrete failure state, not "chaos"
2. Mechanism (the transferable step) — the rule, control, or habit that causes the change
3. After (proof of the same scene) — same subject, improved by the mechanism

Suggested timing for a ~45s vertical teach:

Hooks still need first-frame / under-3-second SaaS clarity; the before act is not a vague mood montage.

Act writing rules

Before

• Name one failure metric or failure moment ("wrong queue," "CSV rework," "double invite")
• Show one visual family you can reuse in After
• Avoid stacking three unrelated pains—pick the pain the mechanism fixes

Mechanism

• One rule or one control cluster
• Captions carry the rule even if VO exists
• If the step needs two sub-steps, split into two mechanism beats—do not compress into a blur

After

When before/after beats process teach

Use before/after when:

• The outcome is visually comparable in one frame family
• The mechanism fits in one rule (or one control cluster)
• Viewers already know the domain enough to recognize "broken"

Prefer process teach (step 1 → 2 → 3 with no dramatic after) when:

• The value is invisible (compliance, latency, trust)
• The "after" looks identical to "before" on screen
• You need more than one mechanism to avoid false simplicity

A useful test: if removing the mechanism act still "makes sense" as a glow-up ad, you do not have a teach Reel yet. Another test: if After cannot reuse Before's crop, you are showing a different scene, not a transformation.

Filled SaaS example: inbox routing

Intent: Teach "label intent before auto-assign."

Notice the after reuses the before composition. That reuse is what makes the mechanism feel causal instead of decorative. For loop-friendly endings after the after-act, see loopable explainer endings.

Alternate fill: billing export

Same spine, different product surface. The structure travels; only the proof crops change.

Board notes that keep regenerations local

On each frame mark:

• act: before | mechanism | after
• visual: UI | B-roll | card
• must-match: which before frame the after must echo
• claim: one sentence the captions must not contradict

When AI video drifts, regenerate only the failing act. Do not re-prompt the whole glow-up. The same discipline as avoiding regeneration-forcing mistakes applies: isolate the change.

Stakeholder reviews should ask: "Can I perform the mechanism from the middle act alone?" If not, the Reel is still an ad. Use storyboard review rituals for non-design stakeholders so feedback targets acts, not vibes.

Caption spine examples

Keep captions short enough to read on mute during each act:

If captions narrate feelings ("so frustrating!") more than rules, you slid back into vanity glow-up territory.

Soft close

Approve the three-act board—before, mechanism, after—before any clip render. If you storyboard first, tools like Scro help lock that spine so transformation frames still teach.