I co-own a small coffee shop in Asheville. We have been open for almost five years. Our Instagram had grown slowly to about 6,800 followers, with a stable engagement rate of about 5 percent. Posts of our morning bake regularly got 300 to 500 likes. Reels of the espresso pulls did even better. It was nothing wild, but it was honest growth and the people in our DMs were actual customers.
Then I made the mistake of spending a weekend on Marketing Twitter. I read 11 threads about how to grow on Instagram in 2026. I made a list of every tactic. I gave myself 30 days to implement everything. Here is what happened to our engagement.
Tactic one: post daily, no exceptions
Every thread said the algorithm rewards consistency. I had been posting three times a week. I went to once a day, every day, including weekends. The first week the new posts averaged 180 likes instead of our usual 350. I told myself the algorithm needed to adjust to the new frequency.
It did not adjust. By week three, posts were averaging 90 likes. The followers who actually engaged were getting fatigued. The algorithm was showing my content to fewer people because the per-post engagement rate had dropped. Posting more had not given me more reach. It had given me less.
Tactic two: jump on every trending audio
I was told that using trending audio is the single best way to get into the Explore page. So I started chasing trends. A trending sound about office workers. A trending sound about toxic relationships. A trending sound about overpriced apartments. I awkwardly tied each one to coffee.
One Reel got 14,000 views, which was three times our previous best. The followers it brought us? Almost none. The reach we got on the next regular post? Half of what it should have been. The algorithm had decided we were now a 'trending audio' account, not a 'local coffee shop' account, and the audience it sent us was the wrong audience.
Tactic three: write 'longer captions with a hook'
The Marketing Twitter people loved long captions. They told me the first line should be a hook. They told me to break captions into short lines. They told me to add a CTA at the end. I tried it.
Every caption I wrote made me cringe. I sounded like a LinkedIn influencer who had never operated a real business.
Engagement on those posts dropped 60 percent compared to our normal short captions. The voice was wrong. Our followers were locals who came in three times a week. They did not want a 'three-part framework for building your morning routine'. They wanted to know if we had a new pastry.
Tactic four: comment-pod-style 'engagement loops'
A thread told me to create reciprocal comment chains with five other local businesses. We would all comment on each other's posts within the first hour to boost early engagement. I set this up. It worked for about a week, then Instagram cracked down on coordinated comment activity. Our reach dropped further. Two of the businesses in the loop got their accounts flagged. Mine did not, but the reach hit took six weeks to recover.
The damage report at day 30
- Average likes per post: 350 → 71.
- Average comments per post: 18 → 4.
- Engagement rate: 5.1% → 1.0%.
- Followers: +120 (mostly low-quality from the viral Reel).
- DMs from real customers per week: 8 → 2.
- Walk-ins who mentioned the Instagram in some way: 12 per week → 3 per week.
Recovery and what actually works for us
It took about three months to undo the damage. I went back to three posts a week, all in our actual voice. I stopped chasing trending audio. I wrote captions the way I would talk to a customer at the bar. I let the engagement rate climb back. By month four we were back to roughly where we had been.
The lesson was not that social media advice is all bad. Some of it is great for someone. The lesson is that almost none of it was right for a small local coffee shop. The advice that grows a personal brand from 5,000 to 50,000 followers will actively damage a neighborhood business. Different goals, different game.