Why video sells, what the Instagram algorithm is really predicting, and whose hand people believe.

You bought something this week you did not need.

Maybe it was a hot sauce, a shirt, a candle that smells like a cabin you have never been to. You could build a spreadsheet to justify it after the fact, but that is not why you bought it. Somewhere between seeing it and paying for it, you pictured yourself with it. A little more relaxed. A little funnier. A little more like the kind of person who owns that thing.

Now picture two ads for the same can of soda.

In the first, the can sits on a white background. Perfect lighting, sweating just enough, logo facing front. In the second, the same can is in somebody's hand at a backyard barbecue. There is smoke off the grill, a dog under the table working the crowd, and someone off camera laughing at a story you only caught the end of.

Same can. One ad sells a product. The other sells an afternoon, and your brain puts you in it for free.

The second ad does two things the first one cannot. It makes you feel something, and it tells you a story. Those two work as a pair. The story is the vehicle, the feeling is the cargo, and the sale is where it gets delivered. Most brands are still paying for the first ad. It is costing them sales, and the data says by how much.

Emotion Moves Sales Volume, Not Just Recall

The easy objection is that feelings are soft. Nice for brand people, useless for anyone with a sales target. The numbers disagree.

Nielsen studied 100 ads across 25 consumer goods brands, scored them with EEG based neuroscience testing, then matched them against real sales data. Ads with above average scores were associated with a 23 percent lift in sales volume over the brand's average ad. Ads that scored below average were tied to sales falling 16 percent.

That gap is the part most brands miss. The boring ad is not neutral. It costs you.

The effect does not stop at the first purchase either. Harvard Business Review research reported that customers who are fully emotionally connected to a brand are, on average, 52 percent more valuable than customers who are merely highly satisfied. Satisfied customers are polite. Connected customers come back, spend more, and stop shopping around.

Buyers will tell you this outright if you ask. In Edelman's global brand trust research, 68 percent of people said it is very or extremely important that brands make them feel good. Not inform them. Not impress them. Make them feel good.

I went deep on the advertising effectiveness research, including why funny beats safe by a wide margin, in what makes a creative ad work. The short version holds up everywhere I look: nobody buys the best argued product. They buy the one attached to a feeling.

A Feeling Needs a Story to Ride In

You cannot hand someone a feeling directly. You have to give them something to follow. A person, a problem, a moment where it could go either way. That is a story, and the brain treats it very differently from a list of features.

Neuroeconomist Paul Zak measured this with blood draws. His lab showed people short narratives on video and tested them before and after. Character driven stories consistently caused the brain to release oxytocin, a hormone involved in social bonding and cooperation, and the amount released was associated with how much people were willing to give to a related charity afterward. The story ended. The generosity showed up after it.

One detail from Zak's work is worth holding onto for later. He found oxytocin is also produced when we are trusted or shown kindness. Stories and kindness run on the same wiring.

Stories also change what people will pay. In 2009, journalist Rob Walker and writer Joshua Glenn bought 100 pieces of thrift store junk, asked writers to invent a short story for each one, and listed them on eBay with the stories attached. Every listing said plainly that the story was made up. The buyers knew the stories were fiction, and the $128.74 worth of junk still sold for $3,612.51. Nothing about the objects changed. The only thing added was a story.

Now imagine the story is true. That is the job of a creator video.

A Photo Shows the Product. A Video Shows the Moment.

Before I made videos with a backyard bar in the frame, I spent 25 years in software. I built a company, sold it, and did a lot of usability testing along the way. That job teaches you one thing fast. What people say about a product and what their face does while they use it are two different reports, and the face is usually the honest one.

You see the half second where someone gets it, or gets confused, or gets delighted. That half second tells you more than an hour of survey answers.

Video can capture that half second in motion and hand it to a stranger. A photo can show a can in a hand. It cannot show the guy holding it pause mid sentence, look at it, and grin like he forgot he had it. It cannot show timing. It cannot carry a laugh. It definitely cannot show the dog stealing a hot dog in the background, which is the most relatable thing that has ever happened at a barbecue.

Short form video compresses a whole story into seconds. My tree videos run on the oldest structure there is. A big tree in suburbia, swaying. The camera goes up. There is a grown man in it with your product, and the viewer has one question: why? Setup, tension, payoff, and the product sits right in the middle of the answer. When I locked a pita in a safe for Pita Mania, the question was the same, and 22,800 people stayed for the answer without a dollar of ad spend.

Stories are also why people stay. By my own numbers, my videos average 110 percent watch time. On a 15 second video, the average viewer watches about 16.5 seconds, which means a lot of people finish and go around again. Nobody rewatches a spec sheet. People rewatch a story they felt something about.

Two Ways to Put a Product in a Story

There are two ways to build a story around a product. Both work, as long as each one is tied to one clear feeling.

In the first, the story is about the product. It is the main character, or the thing everybody in the story wants.

In the second, you tell a story about something else entirely, and the product is simply in view the whole time. In the hand, on the counter, part of the scene. Nobody gets pitched. The viewer feels something, and the feeling attaches to whatever was on screen while they felt it.

To show the difference, I made up a brand. There is no Back Porch Heat hot sauce, so please do not go looking for it. My neighbours already asked.

The story about the product. The feeling is belonging. Sunday barbecue. Three brothers, one bottle left. Each of them hides it from the others, behind the ketchup, inside an oven mitt, under the ice in the cooler. Final shot: their mother pulls the bottle out of her purse, puts it on her burger, and says nothing. The table goes quiet. The sauce is the whole plot, and the feeling is family. Being the thing a family fights over is the highest compliment a condiment can get.

The product in view while the story runs. The feeling is laughter. I am cooking wings in the backyard and telling the story of the worst first date I have ever been on. The bottle is in my hand, going on the wings, sitting on the bar. I never say its name. The punchline lands, I take a bite, my eyes water, and the bottle is sitting in the frame looking innocent. Nobody was sold anything. They laughed with that bottle in view, and that is the association that sticks.

Use the first when the product has a real story of its own: a recipe, a founder, a reason it exists. Use the second when the product is ordinary and the moment is not. Most brands forget the second one exists, and it is the one that feels least like an ad. Both keep the brand name late, which, as the sharing research further down shows, is where it belongs.

One Short Form Story, Second by Second

The wings video, written out as a 15 second reel:

0 to 2 seconds. Tight on me at the grill. "The worst date of my life ordered for me."

2 to 8 seconds. Sauce going on the wings. "She told the waiter I looked like a mild guy."

8 to 12 seconds. I take a bite. Long pause. Eyes watering. I do not break.

12 to 15 seconds. "She was right." The bottle sits in the frame. The name shows up for the first time as text on screen.

Fifteen seconds holds a character, a problem, a turn and a payoff, with the product in view from the first frame and named in the last one. The feeling does the selling. The bottle just has to be there when it happens.

The Long Version Answers the Questions

The short cut sells the feeling. Then the viewer has questions. How hot is it, really? What do I put it on? Where do I buy it? A 15 second reel cannot answer those, and it should not try.

That is the job of long form video: a 3 to 6 minute demo or walk through of the questions buyers ask, chaptered so Google, ChatGPT and YouTube have something substantial to understand and cite. For Back Porch Heat, it would be the full backyard cook, a heat test on a few brave neighbours, and a chapter on what it goes on. The story does not disappear in the long version. It gives the facts somewhere to live, which is why people watch past the first minute.

Same shoot, same bottle, same feeling, two jobs. The short one sells. The long one gives buyers, and the search and AI systems they ask, a real answer to find and cite. I break down how one shoot does both jobs.

The Instagram Algorithm Is Part Technology and Mostly Psychology

People talk about the algorithm like it is weather. Something that happens to you. Instagram has explained a good chunk of how it works, and my read after three and a half years of posting is simple: it is technology built to predict human behaviour, which makes it mostly psychology.

The technology part is prediction. Instagram's own explanation of Reels ranking says the most important things it predicts are whether you will watch a reel all the way through, like it, go to the audio page, and say it was entertaining or funny. Meta's transparency pages describe the feed system estimating how likely you are to skip a post compared with the posts around it.

Look at that list the way a psychologist would. Will you stay. Will you laugh. Will you get bored. Those are questions about people, not code. Somewhere a very large server farm is trying to guess whether you will find a man in a tree funny, and it is spending a lot of electricity on it.

Adam Mosseri, who runs Instagram, narrowed it down in early 2025. He said the three signals that matter most for ranking are watch time, likes, and sends, with likes weighing a little more among your followers and sends weighing more for reaching people who do not follow you yet.

The way I see it, every one of those is a feeling wearing a number. Watch time is curiosity, which is what a story creates. A like is a small yes. A send is the big one.

Feelings Get Sent. Sends Get Reach.

A send means you felt something strong enough that you needed one specific person to feel it too. That makes it the signal where feelings, stories and the algorithm meet, so it is worth looking at what earns one.

The best research on this comes from a USC team that studied what drives sharing of video ads across social platforms, testing 11 emotions and more than 60 ad characteristics in two separate field studies. Ads focused on information were shared significantly less. Ads that triggered amusement, excitement, inspiration or warmth were shared more. And the things that created those emotions were story elements: surprise, plot, and characters, including babies and animals. So the dog under the table is doing real work.

The same study found something every brand manager should hear twice. Prominent brand placement, meaning the name shows up early or stays on screen the whole time, was associated with lower sharing. Yet the researchers found most ads do the opposite, leaning on information, heavy brand placement and very little surprise. The instinct to put the logo on the first frame works against the video travelling.

Wharton researchers studying what people pass along found the same shape from another angle. Content that triggers high energy emotions like awe or amusement was more likely to be shared, while content that left people flat or sad was less likely to be shared, even after they controlled for how useful or surprising it was.

Now put that next to the platform. Mosseri has said messaging, Reels and recommendations have driven most of Instagram's recent growth, and the Mobile Ecosystem Forum's consumer surveys found the share of smartphone users who use Instagram to message friends and family rose from 36 percent in 2022 to 48 percent in 2025. People are not only watching on Instagram. They are talking there, and your video is either part of the conversation or it is not.

So the chain runs like this. A story creates a feeling. The feeling earns a send. The send tells Instagram to show the video to people who have never heard of you. Those strangers watch, feel it, and send it again. None of that is a hack. It is psychology with a delivery truck attached.

To be straight with you, the research backs each link in that chain on its own, not the whole chain as a guarantee. Emotion moves ad response. Stories and emotion drive sharing. Sharing drives distribution. Trust shapes how people respond to a brand. Put those together, measure against real sales, and you have a system worth running. Anyone who promises you the whole chain every time is selling something, and it is not a feeling.

I wrote about where those sends go, and why the sale they start never shows up in your report, in the sale happens in a DM you will never see. The same logic runs discovery now, since Instagram is a search engine and more of the path from first look to purchase happens inside the app, as covered in your buyer journey flipped.

Trust Decides Whose Hand Gets Believed

A feeling only transfers if the viewer believes the person telling the story.

Put that can in a stranger's hand and the viewer's first thought is "ad." Put it in the hand of someone they have watched for years, someone who has made them laugh on a bad day, and the thought becomes "if he likes it, maybe I would too." Same can. Same lighting. Completely different result.

The data backs this hard. Richard Edelman says that for the first time in his firm's research, trust now sits level with price and quality in how people choose brands. Edelman's 2026 research goes further on creators specifically. Among the 48 percent of people who trust a food or lifestyle influencer, 62 percent said they would trust, or consider trusting, a company they currently distrust if that influencer vouched for it. Their brand report found people are nearly twice as willing to back a brand when it has earned both trust and relevance.

This is also why AI video will not save you money. Mosseri said it himself in his New Year's post: authenticity is becoming scarce, which will drive more demand for creator content, and the creators who stay real, transparent and consistent will be the ones who stand out. A synthetic spokesperson can hold the can. Nobody has ever been vouched for by one. I ran the full math in why hire real human reel creators when AI video is cheaper.

Trust also feeds the AI answer engines. When they recommend brands, they draw on far more than your website: reviews, videos, communities, and what other people publish about you. What others say about you is part of what the machine reads. More on that in how ChatGPT decides which brand to recommend.

What I Own After Three and a Half Years

I do not own Instagram. I do not own the algorithm, and it changes its mind more often than I do. What I own is the trust of the people who watch me, and I have spent three and a half years building it one video at a time.

I did not build it with a strategy deck. I built it by being the same guy on camera and off. I live life to the fullest every day, I do not care much what people think, and I try to be kind. The HOA has opinions about my backyard bar. I have made my peace with that.

I even put it on a shirt. The NAPD shirt reads: "Be whatever you want to be, just don't be a friken asshole." That is the whole brand policy. It fits on one line.

Remember Zak's finding that the brain produces oxytocin when someone is trusted or shown kindness. That is the whole reason being decent off camera shows up on camera. People can feel it.

You do not have to take my word for any of this, which is the point. Read the comments.

On one of my videos, Mind Connect Psychotherapy left this: "People are going to judge you no matter what, live your life my friend. Anyone who takes 5 minutes to get to know you will understand you have a big heart." Full disclosure, I have done work with Mind Connect, and I recommend them because I believe in what they do. That goes both directions, which is how trust is supposed to work.

On "The last thing my ex said to me at this bar", a viewer wrote that I was an amazing human, called out the energy, the humour, and the Crocs, and told me to keep being me. That title is a whole story in eleven words: a character, a history, and a line you have to hear. I had also met her in person at a run. That is the part a brand should notice. The people in my comments are not an audience I rented. A lot of them are neighbours, people I have met, people who would tell me if I started selling them garbage.

Which is why I turn brands down. If I would not use it, I do not post it. That is not a moral flex. It is the performance mechanism. The day my audience catches me faking enthusiasm, the story stops being true, the feeling goes with it, and a brand paying me after that is paying for nothing. Any creator you hire should be able to show you what they said no to. I explain how to check in how to vet a UGC creator portfolio.

Feeling Is the Mechanism. Sales Are the Scoreboard.

None of this means chasing views. I get about four million views a month, and I will be the first to tell you reach is inventory, not a result. The goal is almost always to sell more. Sometimes it is sign ups, bookings, or people walking through a door. Name the goal first, then measure against it.

Brand lift sits in front of that. Ad recall, consideration, purchase intent. It is the early sign that the feeling landed. Sends per reach belongs on that list too, because it tells you whether the story was good enough to pass along. All of it feeds the sale. None of it is the sale.

Feeling gets tested like anything else. I start organic, small and cheap, with no ad budget. I find what works, change one thing, and run it again to beat my own winner before a paid dollar goes in. One round of testing is a single data point pretending to be a pattern. When this is done right, the money side changes fast, as the numbers in short form video cut ad spend 85 percent show. If you want the wider evidence on creator content as a sales channel, it is all in does UGC actually work.

Questions to Ask Anyone You Hire to Make People Feel Something

• What feeling is this video selling, in one word?
• What is the story, and where does the product sit in it?
• How late does the logo show up?
• Would someone send it to a friend, and why?
• Why does your audience believe you?
• Can I read your comments?
• What have you turned down?
• How will we know it sold, not just that it was seen?
• What do we change and rerun if the first version misses?

I answer every one of those before you pay. If that is the kind of creator you want holding your product, let's talk.

Your can on a white background is just a can. Put it in the right hand, give it a story, and people want to be there.