The Instagram Algorithm Is A Moving Target, But It’s Interesting

If you actively post on a social network, you want that post to be seen by as many people as possible. As we know, that’s getting more difficult, especially on Instagram. Stories abound about creators finally achieving hard-earned virality only to have IG change its algorithm, which then takes the user’s reach back to almost nothing. Podswap has a section of its site where it reverse engineers the Instagram algorithm to find out what’s working and what’s not. As you’ll see, that’s a moving target that changes every month.

The Metrics That Count

To start, Podswap states, “2026 marks the official consolidation of the ‘Relational Economy’ update. The algorithm has shifted entirely away from hollow engagement signals (likes, views) towards high-friction interactions that signal genuine human connection. ‘Bridge Content’ posts that are shared across disconnected friend groups are now the primary driver of viral discovery.”

The IG algorithm has changed so the metrics have changed with it. As a result:

• Shares are now weighted 4x higher than likes. Content must be designed to be sent, not just consumed. The ‘Share to DM’ velocity is the strongest viral signal.

• AI detection for ‘perfect’ commercial aesthetics has tightened. Raw, phone-shot visuals are outperforming studio-produced content by 35% in discovery feeds.

• Saves are the new ‘Super Like’. They indicate utility and intent to return, signaling long-term value to the algorithm, whereas likes are now considered a ‘passive’ signal.

• Keyword optimization resulted in a direct +35.6% lift in discovery reach across all account sizes.

Some other insights for April:

• The best performance came when using 2 hashtags

• A caption length of 100 to 149 words received the most action

• A statement post worked better than asking a question

• The sweet spot for posting was once a day

• The best day of the week to post was Monday and the worst was Tuesday

• The best times to post were 3PM, 2PM and 9AM

• Adding a tagged location provided a 215% lift in reach

Best Performing Niches

By far the most variable data points from month to month were the best and worst performing niches. For instance best performing in the month of April were:

1. AI Art Generation Tools & Techniques
2. Photo & Video Editing Software
3. Photography & Videography Gear & Technology
4. Stop Motion Animation (Art)
5. Digital Art & Illustration

While worst performing were:

1. Tech News, Trends & Analysis
2. Book Collecting & Rare Books
3. Small Businesses & Local Entrepreneurs
4. AI in Business & Automation Software
5. AI & Machine Learning (General)

Of note are the worst performing niches, with Tech News being worst, while AI-related niches are at 4 and 5.

Now let’s contrast that with the best performing niches in February:

1. AI in Business & Automation Software
2. AI & Machine Learning (General)

The best performing niches in one month were the worst performing in another.

Also different in February were the best caption length (400+ words) and best hashtag density (5 hashtags).

Be Careful Of The Data

What all this means is that you have to be careful about whatever data sources you might be following. Trying to optimize your posts using data that is only a month or two old might take you in the wrong direction and result in diminished reach.

Social algorithms are changing more frequently than, so they’re truly a moving target.

Trying to optimize a post could result in spending a lot of time on something that used to work but doesn’t currently.

The best direction might be following what Google suggests for optimizing websites, which is to create a post with the best content possible that adds real value for the reader. And then cross your fingers.

---