The (Social Media)
Theory of
Everything
742 vertical videos. Instagram Reels + TikTok. 5 sectors.
One pattern that never changes.
We make films
that move people.
We ran this study
to understand exactly why.
Black Iris Films is an emotional storytelling studio. We've spent years watching the same pattern play out: the content that actually travels — the videos people share, save, and show a colleague — is almost always emotionally-led.
We wanted to know if the data agreed. So we analysed 742 short-form videos across five sectors, 48 brands, and two platforms. This is what we found.
What you'll learn
Five stops on the road map. Each one a finding that changes how you brief, measure, and pitch emotional content.
We already know
emotion drives revenue.
Here's what no one
had measured.
The business case for emotional marketing is established. Harvard Business Review found emotionally connected customers are worth 52% more over their lifetime than customers who are merely satisfied. The IPA's 2025 analysis found emotion-rich creative is roughly twice as likely to deliver large profit gains as purely rational advertising.
What none of that tells you is what happens before a customer has even heard of your brand. Specifically: does emotion-led short-form video drive the social engagement that builds brand affinity in the first place?
That's what this study set out to answer. If emotion drives purchase decisions, and social media determines whether your brand makes someone's shortlist at all, then whether emotion works on Reels and TikTok isn't academic. It's the whole game.
So we ran the study. Here's what we found.
more lifetime value from emotionally connected customers vs. merely satisfied ones
Harvard Business Review
more likely to deliver large profit gains — emotion-rich campaigns vs. rational ads
IPA, 2025
"The video you spent three weeks on did nothing.
The one someone threw together on a Tuesday afternoon went semi-viral.
You blamed the algorithm."
It wasn't the algorithm.
The one thing
that never changes
Every marketing team has chased the algorithm. Posted at the right time, used the right audio, followed the trend cycle. All real. All temporary.
Underneath all of it, there's a constant — something true before TikTok existed and true after whatever replaces it. The videos that actually travel are almost never the informational ones. We kept seeing it ourselves. So we ran the study across 742 short-form videos and five sectors to either prove it or kill it.
The study
We analysed 742 short-form vertical videos — exclusively Instagram Reels and TikTok — across five sectors: financial services, fintech, AI and B2B tech, higher education, and a consumer/lifestyle control group.
Every video was tagged as either emotion-led or informational, then sorted into seven emotional categories: empathy, pride, aspiration, authority/trust, humour, fear, and curiosity. Videos were then ranked by how well they performed.
The result is the most detailed dataset we've found on what drives short-form video engagement in sectors where buying takes real thought. As far as we could tell, no one else had run it.
One thing worth saying upfront: we studied vertical video because it gives you the most transparent feedback loop in media right now. Millions of people making a split-second decision on whether something is worth their attention — no loyalty, no obligation, no social contract with the brand. The signal is immediate and the sample is enormous.
But what the data reveals isn't really a vertical video finding. It's a human one. The content that travels has always been the content that makes people feel something — whether that's a 10-second Reel or a three-minute brand film. The format changes. The psychology doesn't.
Emotion doesn't raise
the average.
It raises the ceiling.
This is the most important thing to understand about the data — and the most misunderstood thing about emotional content.
Does it actually move the needle?
Emotional posts are
2.1× more likely
to break out
Here's the finding people get wrong first. Emotional content doesn't reliably raise the average. Across target sectors, the median engagement uplift for emotional content is +5.0%. Real, but not a mandate.
The median is not the story. The ceiling is.
Think of it like basketball: emotion doesn't make your whole team taller. It makes your best player taller. You're not optimising every video — you're setting up the conditions for the one that takes off.
Emotional posts hit the top quartile 28.4% of the time, versus 16.0% for informational posts. After accounting for sector, platform, production quality, and video length, that's 2.1× the odds of breaking out.
Stop measuring emotional content against the same benchmark as informational. Give it a separate success metric: ceiling performance, not average lift.
Emotional posts are nearly twice as likely to break into the top quartile of performance.
the odds of breaking out, controlling for sector, platform, production quality, and video length.
27 of 33 brands.
Same result.
We looked at the 33 brands in the dataset that posted both emotional and informational content — where we could compare like for like.
Emotional posts beat informational on median performance in 14 of 33 brands. Unremarkable. But they beat informational on maximum performance in 27 of 33 brands. That's not noise. That's the whole point, right there.
When pitching emotional content internally, lead with this: 27 of 33 brands, same result. It's the clearest version of the argument.
Emotion doesn't consistently win the average. It almost always wins the peak.
More upside.
No meaningful
downside.
Emotional content is not a high-risk bet. This is not a "higher ceiling, lower floor" trade-off.
In the sectors we studied, emotional posts make up 12.7% of the top 10% of all content. Informational posts: 2.7%. At the bottom 10% — the worst-performing content — emotional posts are 9.6% versus 11.3% for informational. Emotional content slightly outperforms informational even among the weakest posts.
More upside. No meaningful downside. That's as close to a free lunch as marketing data gives you.
This data kills the "it's too risky" objection. Emotional content doesn't fail harder when it misses. It just wins more when it lands.
The distribution is wider for emotional content — more upside, similar floor.
One sector makes the case
so cleanly
it almost feels unfair.
What does it look like when a sector fully leans in?
Financial services.
+114% median uplift.
Both platforms. No caveats.
Financial services isn't just the strongest sector in this study — it's the only one where the case for emotional content is strong enough to lead a pitch with full confidence.
Emotional finance posts generate a +114% median engagement uplift compared to informational content. They hit the top quartile 30.8% of the time, versus just 8.3% for informational posts. And while emotional content makes up 76.5% of all finance posts in the dataset, it accounts for 92.3% of the top quartile. Emotion already dominates the space — and it still punches above its weight at the very top.
Why does finance respond so strongly? Because money is personal. People lie awake at night about it. They feel shame, pride, fear, and ambition around their finances — often in the same week. Finance was always an emotional category. Most brands just forgot.
If you're in financial services and your content calendar is still built around rates, features, and product announcements, your competitors' emotional content is already claiming the ceiling you're leaving empty.
Median engagement uplift for emotional vs informational finance content
Top-quartile hit rate: emotional vs informational
of the top-quartile finance posts are emotion-led
Positive on both platforms — the only sector where this holds.
What if my sector isn't finance?
Every sector.
Same pattern
at the top.
Fair question. If you're not in finance, you're probably looking at that +114% figure and wondering whether any of this applies to you. Short answer: yes, but differently. In fintech, the median uplift for emotional content is actually negative (-15.8%). In higher education: -5.6%. B2B and AI: near neutral.
On the average numbers alone, you could argue emotion doesn't reliably work in those sectors. But look at the top of the pack. In fintech, 75% of the top-quartile posts are emotion-led. In higher education, 85%. Even where emotion doesn't lift the average, it still owns the peaks.
The pattern is the same everywhere. What varies is how reliably emotion lifts the average — not whether it dominates when things go right.
The right benchmark isn't "does emotional content beat informational on average in my sector?" It's "what does the best content in my sector actually look like?" The answer is always emotional.
| Sector | Median uplift | Top-quartile % emotional |
|---|---|---|
| Financial Services | +114% | 92% |
| Higher Education | −5.6% | 85% |
| Fintech | −15.8% | 75% |
| AI / B2B Tech | +5.1% | 62% |
Median varies. The top is always emotional.
% of top-quartile posts that are emotion-led
Should we be doing this on Instagram or TikTok?
Where you post
changes everything.
The sector averages hide a platform split that should change how you brief creative — especially in B2B. Finance is the exception: positive on both platforms. Everywhere else, the platform matters as much as the sector.
For B2B and AI brands: the honest story isn't "emotion works." It's "emotion is a TikTok play." Running emotional B2B content on Instagram and wondering why it underperforms? The data has your answer.
Brief your emotional content by platform, not just by sector. The same video strategy doesn't apply across both channels. For B2B, the TikTok brief and the Instagram brief should look different.
| Sector | TikTok | |
|---|---|---|
| Financial Services | +88% | +159% |
| AI / B2B Tech | −24% | +62% |
| Fintech | Slightly negative | Slightly negative |
| Higher Education | ~ Flat | Negative |
Finance is the only sector that works on both platforms. For B2B, emotion is a TikTok strategy.
Not all emotion is
created equal.
If your brand's answer to "more emotion" is "let's do a funny video," this section is for you.
Which emotions
should actually be
in your brief?
If you asked most marketing teams what kind of emotional content performs best, they'd probably say humour — or something that grabs attention. The data disagrees. There are seven emotion types in this study, and the gap between the ones that work and the ones that don't is bigger than you'd expect.
The winning family — empathy, pride, belonging, aspiration, and trust — hits the top quartile 39% of the time. The losing family — humour, fear, and curiosity — hits it 16.7% of the time. After accounting for brand differences, the winning group is 3.3× more likely to break out (p<0.001).
These emotions work because they mirror what someone feels during a big financial or career decision: recognition, a sense of identity, confidence about the future. They go somewhere real. The losing emotions — however attention-grabbing — tend to stay on the surface.
The emotion should be named in the brief before a single frame is storyboarded. Not "emotional content" — which emotion, specifically. Empathy, belonging, aspiration, or trust. That decision belongs upstream, not in the edit.
* Directional only — based on n=20
Two things the data says that your instincts probably don't.
Fear doesn't work.
Think about how many finance ads open with some version of "most Australians don't have enough saved for retirement." You know the ones. That approach is one of the worst-performing emotion types in the entire study.
The audience can feel when you're trying to scare them into engaging. They scroll.
Curiosity doesn't work either.
Short-form platforms are obsessed with pattern-interrupt hooks — surprising statements, cliffhangers, unexpected openings. Hook strength alone isn't enough.
What travels in these sectors isn't novelty. It's identity, belonging, and trust. Content that makes someone feel seen outperforms content that makes them feel surprised.
Audit your current hooks. If most of them open with a risk statement or a "did you know" — the two lowest-performing emotion types in the study — that's your starting point.
The wild card:
humour is a
half-court shot.
Here's where it gets confusing. Humour is technically in the losing group — 21% top-quartile hit rate, well below empathy and pride. And yet 40% of the top 10 videos in this dataset are humorous. Both of those things are true at the same time. That's not a contradiction. That's the definition of high ceiling, low floor. Humour can produce the single best video your brand has ever made. It can also produce a lot of content that quietly goes nowhere, dragging your average down while delivering the occasional spike that everyone in the meeting claims credit for.
The rule: take the half-court shot sometimes. It's absolutely worth attempting. Just don't build your offence around it.
Humour should be a deliberate creative swing, not your default emotional register. Reserve it for moments when you have the creative confidence and production quality to land it. Then go all in.
Overall top-quartile hit rate for humour
Share of the top 10 posts that are humorous
High ceiling. Inconsistent floor. Use deliberately, not as a default.
Forget the averages
for a second.
Here's what the
actual best posts look like.
The top 10 cluster into two types.
Stop looking at the averages. Start looking at individual videos. The top performers in this dataset don't look random — they cluster around two distinct archetypes. Understanding them is more useful than any single statistic.
- Playful, immediate, often participatory
- Designed to trigger a social reflex — tag someone, comment, share
- Minimal brand presence
- No product explanation
- Hook is simple and direct
- Cultural cachet, visual quality, status-adjacent
- Creates an identity the viewer wants to associate with
- Confidence and aspiration are the primary mechanisms
- Brand presence is present but not dominant
Which archetype fits your brand right now? Social-native content can be built quickly and tested cheaply. High-production aspirational content raises the stakes — and the ceiling. Both work. Defaulting to neither is the problem.
What unites
every top performer
Feeling.
Top content doesn't try to communicate clearly. It tries to create a feeling, instantly. That is a fundamentally different creative brief.
Hooks in the top 10 are simple and direct — not complex setups, not questions. They start in the middle of something.
mechanics.
A meaningful number of top performers are built around commenting, tagging, giveaways, or community interaction. Built to get people to do something, not just watch.
or absent.
Product explanation is not the point. The feeling is the point. The brand follows.
Most top performers are 6–16 seconds. The best posts don't waste a frame.
The framework.
Everything above was the why. Here's the how.
Three steps.
Every emotional video.
This framework comes from what we kept seeing in the data — and from something screenwriters have always known. In film, every story makes a promise to the audience. If it doesn't deliver on that promise, the audience feels cheated. Short-form video works exactly the same way. Videos that accumulated high views but low engagement almost always failed at the same point: they signalled an emotion they never followed through on. FEEL → SHOW → DELIVER is how you fix it.
Decide the specific emotion you want to evoke before you brief a single frame. Not "emotional content" — which emotion. Empathy, belonging, aspiration, and trust are what works in sectors like yours. Fear and curiosity are not. This decision belongs in the brief.
Signal that emotion in the first three seconds. Top performers don't build to a feeling — they start there. The hook is a promise: this video will make you feel something specific.
Storytelling hooks outperform question hooks: 42% vs 20.3% top-quartile rate. Start in the middle of something real. Not with a question.
Follow through on the emotional promise you made in the first three seconds. This is where most videos fail — they hook with emotion, then pivot to a product message that breaks the contract entirely.
Emotional videos without a CTA hit the top quartile 31% of the time, versus 21.2% with a CTA. Adding a CTA roughly halves top-quartile odds for emotional content.
Take your last three emotional videos. Did they promise a feeling in the first three seconds and actually deliver it? Or did they hook emotionally and finish as product content? That gap is where most brands lose their ceiling.
Five rules
from the data
These aren't conventions — they're findings. Each one comes directly from the dataset.
16–30 seconds is the strongest overall band. The top 10 skew shorter: 6–16 seconds. The best posts don't waste a frame.
Use storytelling hooks, not question hooks. Storytelling hits the top quartile 42% of the time. Questions: 20.3%. Start in the middle of something real.
Leave it off emotional posts if you're aiming for high engagement. A CTA roughly halves top-quartile odds for emotional content. Run CTAs on informational posts. Let emotional content breathe.
Production quality doesn't predict breakout engagement on its own — but that's not a reason to go lo-fi. High production signals credibility and earns trust. The story underneath still has to do the work. Polish without emotional craft is expensive content that goes nowhere.
Before you brief a single shot, decide what you want people to do with the video — comment, tag a friend, enter a giveaway. Build that intention into the creative from the start, not as an afterthought in the edit. The algorithm rewards action, not just views.
What the data
can — and can't —
tell us
Before you go brief a bunch of empathy videos — a few honest things this study can and can't tell you.
Authority/Trust is a promising emotion type at 30% top-quartile hit rate — but it's based on n=20 videos in the target sectors. Treat it as directional until the sample grows. Fintech is genuinely ambiguous as an always-on emotional strategy — it works for peaks, not baselines. Engagement in this study captures likes and comments per view; shares and saves data is incomplete, which means cross-platform absolute comparisons carry some caveats. And this study measures platform engagement — not revenue, not pipeline, not brand recall. The link to business outcomes is directional, not causal.
One finding worth noting on the other side: when the analysis is restricted to the clearest, most unambiguous classifications only, the average advantage for emotional content rises from +5.0% to +26.9%. The true effect is likely stronger than the headline number, not weaker.
The broad emotional bucket includes some borderline cases.
When those are removed, the advantage for emotional content nearly triples. The headline number is conservative, not generous.
Key takeaways
What the data says
The headline findings.
What to do differently
The playbook.
The algorithm
will keep changing.
This won't.
Trends change. Formats change. The algorithm changes. What doesn't change is that humans are emotional beings who share things that make them feel something. That was true before TikTok. It will be true after whatever replaces it. The data in this study is a snapshot of 2024–2026. The pattern it describes is considerably older.
Curious what this
looks like for
your brand?
Take the Black Iris Fit Quiz. Seven questions, one minute, zero commitment.
Tell it what you're trying to achieve, and it'll recommend the right video types, plus a price range so you know what you're working with.
Methodology
About the study
This study was designed and conducted by Black Iris Films to test whether emotional content outperforms informational content across short-form video in considered-purchase sectors. It is the first systematic study of this question we are aware of across these specific verticals.
Data collection
Videos were collected using a hybrid approach: automated extraction via Apify (a web scraping and automation platform) combined with manual review and classification. Each video was captured across an 18-field data template covering post URL, brand, sector, platform, post date, content classification (E/I), emotion subtype, hook type, CTA presence, production quality tier, duration, and engagement metrics (likes, comments, views).
Sample
774 posts across 48 brands and 5 sectors. 32 posts were excluded due to data integrity issues in the likes column, yielding a final analysis base of 742 posts. The target sectors — Financial Services, Fintech, AI & B2B Tech, and Higher Education — were selected for their relevance to considered-purchase decisions. Consumer & Lifestyle was included as a control group. The brand sample is weighted toward Australia, reflecting Black Iris Films' primary market, and structured as follows:
| Origin | Count | Description |
|---|---|---|
| Australian-headquartered | 30 brands | Core sample |
| Global brands with AU presence | 10 brands | Regional relevance |
| Global benchmark brands | 10 brands | International comparison |
Brands were selected for active Instagram Reels and/or TikTok presence and a minimum posting frequency sufficient for meaningful analysis.
Platforms
Instagram Reels and TikTok. LinkedIn was excluded from this iteration due to fundamentally different engagement mechanics and audience behaviour. Platform-specific quartiles were used for cross-platform breakout analysis.
Engagement metric
Engagement was calculated as (likes + comments) / views. Note: shares and saves data was incomplete across the full dataset — particularly for Instagram — so the metric reflects the consistently available fields. Cross-platform absolute comparisons should be read directionally rather than precisely.
Performance tiers
Posts were ranked by engagement per view and divided into quartiles. The top-quartile hit rate — the percentage of posts reaching the top 25% of performers — is the primary metric throughout this report. Note: the original study design specified tertiles; quartile analysis was adopted during execution as a more granular and commercially intuitive frame.
Confidence scoring
Every classification was assigned a confidence score (High / Medium / Low) based on the clarity of the emotional or informational signal. Low-confidence posts — typically hybrid content where the emotional hook was ambiguous — are included in the headline analysis but flagged in sensitivity analysis. When the analysis is restricted to high-confidence classifications only, the emotional median advantage rises from +5.0% to +26.9%.
Inter-rater reliability
The study design targets a Cohen's Kappa of ≥0.70 on a 10% subsample independently coded by a second rater. Full kappa results will be reported in Phase 2 methodology documentation.
Statistical analysis
Top-quartile comparisons use chi-square tests. Within-brand peak analysis uses a binomial test. The controlled model is a brand-clustered logistic regression with sector, platform, production quality tier, and post duration as covariates. All reported p-values are two-tailed.
Classification rubric
The core question
Every video was classified by asking a single question in the first three seconds of viewing:
"Was my first instinct to FEEL something — or to LEARN something?"
This is the 3-second rule. It reflects how audiences actually engage with content: the hook determines whether the brain enters an emotional processing mode or an analytical one. Videos where the first three seconds trigger a feeling — connection, pride, aspiration, humour, fear, curiosity — are classified Emotion-Led (E). Videos where the first three seconds prompt an informational response — "tell me more," "I didn't know that," "that's useful" — are classified Informational (I).
Hybrid videos
When a video contains both emotional and informational elements, it is classified by whichever element dominates the hook. If the hook remains ambiguous after careful review, the classification defaults to Emotion-Led and receives a Low confidence score.
Production quality tiers
| Tier | Description |
|---|---|
| Platform-native | Shot on phone, minimal editing, lo-fi aesthetic — intentionally native to the platform |
| Standard Professional | Competently produced, branded, clear audio/visuals — typical agency or in-house output |
| High Production | Cinematic treatment, dedicated crew, post-production, broadcast-level finish |
Note: production quality tier was not a meaningful predictor of top-quartile engagement performance once sector, platform, and content class were controlled for. This finding speaks to what drives algorithmic breakout — not to the broader value of high production, which remains important for brand credibility, audience trust, and contextual appropriateness.
The 7 emotion subtypes
The video makes you feel seen, understood, or emotionally recognised. Common in: customer stories, relief narratives, community moments.
The video taps into identity — group membership, achievement, shared values, "we are the kind of people who…". Common in: brand community content, milestone moments, values-led content.
The video makes you want something — a version of yourself, a lifestyle, a future state — or creates a sense of wonder. Common in: aspirational lifestyle content, transformation stories.
The video builds credibility and confidence. The dominant feeling is reassurance or respect. Common in: expert positioning, credentials, proof-of-concept content.
*n=20 — treat as directional
The video's primary mechanism is entertainment, levity, or delight. Common in: memes, self-aware brand content, playful formats.
The video activates risk awareness, scarcity, or loss aversion. Common in: cautionary content, deadline-driven content, problem-agitation hooks.
The video's hook is a pattern interrupt — something unexpected, counterintuitive, or "wait, really?". Common in: stat-forward hooks, myth-busting content, surprising reveals.