I own a nine-seat coffee shop in a college town. We do about 180 transactions on a normal weekday and roughly 320 on a busy Saturday. Two years in, we were profitable, the rent was sane, the staff was good. The one thing I could not crack was acquisition. New customers walked in maybe once a week from search or word of mouth, and that was it.
Every twenty-something who walked in had TikTok open. Every cafe within three miles that had blown up in the last year had done it through TikTok. I told my wife I was going to give it 90 days and 90 videos. She said good luck and that I owed her dinner if I failed.
Day 1 to day 12: I almost quit
I posted my first video on a Monday. It was a slow-motion pour of an iced latte, classic stuff. I had watched maybe 50 TikToks from coffee shops to study the format. The video got 89 views. My third post, two days later, got 412. I told myself fine, this is the slow start everyone talks about.
By day 12 I had posted 14 videos. The average view count was 287. My highest was 1,200. None of them turned into a single new walk-in that I could trace. I was waking up at 5am to film latte art, going home after a 12-hour day to edit, and the only return was a slight increase in followers from people who lived nowhere near my shop.
I sat on my couch on day 12 and almost wrote a post telling my followers I was done. The thing that stopped me was realizing I had been making the wrong videos.
What I had been doing wrong
Every video was a beauty shot of a drink. I was filming what I, the owner, thought looked nice. None of the videos had a story, a hook, a problem, a stake, or a face. I was selling vibes to an algorithm that does not care about vibes.
I went back and watched the top 30 cafe videos in my city by view count. Almost none of them were drink shots. They were people. They were arguments. They were behind-the-counter moments. They were the owner getting yelled at by a customer. They were the barista's hand visibly shaking on the steam wand on a Saturday rush. Drama, even tiny low-stakes drama, beats beauty every time.
Day 13 to day 30: I started filming the truth
I made three rules for myself. Rule one: every video has a person in it within the first second. Rule two: every video has a real moment, not a staged one. Rule three: every caption asks a question.
The first video under the new rules was 19 seconds of my morning barista, Jamie, telling the camera why she hates oat milk. Off the cuff, no script. 14,200 views. Two days later I posted a video of me asking a regular what he would change about the shop. He said the music. The video did 38,000 views. People commented suggesting their own playlists. We changed the playlist.
The video that broke through
On day 41 I posted a video of a customer who had been coming in for two years. She walked in, ordered her usual, and I asked the camera if anyone could guess what it was before she said it. Three of my regulars chimed in from off-screen. She ordered her oat-milk cortado with two raw sugars. The video did 410,000 views.
The Saturday after that video posted, we did 412 transactions. Our previous record was 327. The line was out the door from 9am to 1pm. I had to call in a barista on her day off. Three different customers told me they had driven from out of town because of the video.
Day 41 to day 60: the comedown
The viral video taught me a brutal lesson. The next 12 videos I posted all averaged under 5,000 views. The bump in foot traffic faded by week three. People had come once, taken a photo, posted it themselves, and moved on. The algorithm had given me a gift and I had no system to capture the people it sent me.
I started thinking about what I would have done differently if I could replay that week. I would have had a QR code on every cup pointing to a perk. Something like: post a story tagging us, get $2 off your next drink. Something to convert the one-time visitor into a returning one. By the time I built that system, the wave was already over.
Day 60 to day 90: building the system I needed
I spent the last 30 days of the experiment building what I should have built first. A perk on every cup. A simple ask: come back within two weeks for a free drip refill if you bring a friend who has never been here. Half of the customers I brought in from TikTok came back. Most brought someone.
Final numbers
- Videos posted: 87 in 90 days.
- Total views: 1.4 million.
- Followers gained: 14,800.
- Highest single video: 410,000 views.
- New customers I can directly trace: ~340.
- Repeat visits from those new customers: ~190 (56 percent return rate after I added the perk).
- Estimated revenue lift in 90 days: ~$8,200.
- Estimated hours spent filming, editing, posting, and replying: 110.
What I would tell another small coffee shop owner
TikTok works for coffee shops, but not the way Instagram works. Beauty shots are a tax. Real moments are the product. The viral video will come if you keep showing up, but if you have no system to capture the wave when it hits, you will get one good weekend and nothing else.
Soft note: I now use Social Perks to run this. The 14-day free trial was enough time to know it was the system I had been duct-taping together for years.