Most business owners I meet judge a post by its likes. Two hundred likes feels like success. Twenty feels like failure.
The platforms stopped thinking that way a long time ago. That gap is why so many brands post every single day and still don’t grow. Here’s how I explain social media algorithms to my own team at Bizyocon, without the jargon.
What does a social media algorithm actually look for?
An algorithm is a prediction machine. For every post and every person, it asks one question: will this person spend time with this and do something meaningful with it? The posts that get a “yes” from the most people get shown to more people.
Instagram has been unusually open about this. In January 2025, Instagram’s head Adam Mosseri said the three signals that matter most for ranking are watch time, likes, and sends. He suggested creators watch average watch time, likes per reach, and sends per reach (Social Media Today).
Notice two things. Likes are on the list, but measured per person reached, not as a total. And sends sit right next to them.
Why is a like the weakest thing someone can do?
A like costs nothing. You can like a post in half a second without reading it.
A send is different. Someone thinks “my friend needs to see this” and puts their own name behind your content. Mosseri said likes matter slightly more with people who already follow you, while sends matter more for reaching people who don’t. For a business that wants new customers, that’s the whole game.
I think of engagement as a ladder of effort:
- Like: the lowest effort
- Comment: they had something to say
- Save: they want to come back to it
- Share or send: they’re recommending you to someone they trust
So when we plan content now, the question isn’t “will people like this?” It’s “who would someone send this to, and why?”
How does a post reach people who don’t follow you?
Your post isn’t shown to everyone at once. It goes to a small group first, usually a slice of your followers, and the platform watches what they do. If that group stops, watches, saves, and shares, the post is shown to a wider group. Eventually it reaches people who have never heard of you.
Instagram has even turned this into a feature: trial Reels are shown to non-followers first, so you can test an idea without it going to your existing audience.
Two practical lessons follow. The first hour after posting matters, so be around to reply to comments. And a weak post doesn’t just underperform; it tells the platform your content is easy to skip.
What decides watch time on a Reel?
The first three seconds. Your Reel may be thirty seconds long, but for most viewers it’s three, because that’s when they decide to stay or scroll.
Before we touch colours, music, or editing, we check the hook. Five types work reliably:
- Problem: name the pain your customer feels (“Getting clicks but no calls?”)
- Contrast: before and after, or myth and reality
- Number: “3 mistakes that…”
- Curiosity gap: open a question the video answers
- Pattern break: something visually unexpected in the first frame
After the hook, look at the retention graph in your insights. Where viewers drop, fix three things: the pace of the cuts, when text appears on screen, and the ending. A clear payoff or a loop keeps people till the last second.
Is the LinkedIn algorithm different?
The goal is the same: show people posts they’ll find worth their time. The signals look a little different.
LinkedIn’s engineering team has written about using dwell time, how long someone actually spends on a post, as a ranking input. So a post people stop and read can do well even if they don’t react to it.
Comments matter more than reactions, especially thoughtful ones that start a discussion. A generic “Great post!” adds little. A real conversation between people in your industry adds a lot.
Two habits that work for founders on LinkedIn:
- Stay on one topic area, so both the platform and your audience know what you’re an expert in.
- Reply to comments in the first hour. Conversation is a signal. Silence isn’t.
Which format should you use for which goal?
The mistake I see most often: every brief becomes a Reel. Formats do different jobs. This is the rule of thumb we use:
| Format | Best for |
|---|---|
| Reel or short video | Reaching new people |
| Carousel or document post | Saves and depth, when you’re teaching something |
| Static image | Recall, keeping your brand top of mind |
| Story | Trust and DMs with people who already follow you |
Decide the goal first. Then pick the format.
Should you post more often?
Clients ask me this every month: “Can we post more?” My honest answer is usually no.
Each post is judged on its own. Thirty average posts don’t add up to one great one. They spread your audience’s attention thinner and teach the algorithm that your content is skippable. Twelve strong posts a month, each with a clear hook and a reason to share, will almost always do more.
The same thinking applies to selling. Your Instagram isn’t a sales counter. A healthy mix for most businesses is roughly 40% teaching, 30% entertaining or behind-the-scenes, 20% proof like reviews and results, and 10% direct selling.
What should a business owner track?
Likes and follower counts are the tip of the iceberg. The numbers below the waterline tell you far more:
- Sends and shares per reach: is your content worth passing on?
- Average watch time: are people staying?
- Saves: is it useful enough to keep?
- Profile visits and DMs: is content turning into conversations?
- Enquiries: the number that pays salaries
If these are moving in the right direction, likes take care of themselves. I’ve written more about this in Rank #1, zero calls: the marketing metrics that actually matter.
The bottom line
The algorithm isn’t against you. It’s trying to show people things they’ll value. Make posts that people stop for, finish, and want to send to someone, and the algorithm becomes your distribution team.
If you’d like a second pair of eyes on your social media, book a free 30-minute consultation, or see how my team at Bizyocon runs social media for brands across India.
