Instagram chat statistics explained

When you run an Instagram export through Stats Insta, you get a page full of charts. They look simple, but each one is a specific measurement with specific limits. This guide defines every core statistic, explains how it is calculated from your message_1.json file, and — just as importantly — explains what it cannot tell you.

How to see your Instagram chat statistics

Instagram itself shows you almost none of this: the app has no message totals, no per-person breakdowns, and no activity patterns for your own chats. To see your statistics you need your data export and a few minutes:

1. Open Instagram and go to your profile, then the menu, then Accounts Center. Open Your information and permissions and choose Download your information.
2. Request a download of Messages only, set the format to JSON and the date range to All time, then submit the request.
3. When Meta emails you that your file is ready, download the ZIP archive and extract it. Inside, open the messages/inbox folder and find the conversation you want.
4. Select that conversation's files — message_1.json, plus message_2.json and any further numbered parts of the same chat.
5. Open the analyzer, drop the files in, and your statistics appear in seconds, calculated entirely in your browser. The DM export guide covers each of these steps in more detail.

Message counts and message share

The most basic statistic is also the most misread. A message count is the number of message objects in the export attributed to each participant. Message share is that count divided by the conversation total, expressed as a percentage.

Two details matter. First, Instagram counts every send separately: writing one long paragraph and pressing send once is one message, while sending the same text as five short lines is five messages. People who text in bursts will always out-count people who write paragraphs, even if they say less. Second, word and character counts correct for this — if one person has 40 percent of the messages but only 25 percent of the words, they are sending shorter, more frequent texts. Read the three numbers together before deciding who "talks more."

Conversation starters

Replying is easy; starting takes initiative. Stats Insta credits a conversation starter when someone sends a message after the chat has been silent for more than 12 hours. Over a long friendship this produces one of the most revealing numbers in the report: it shows who keeps the connection alive.

A lopsided starter count does not automatically mean the other person does not care — work schedules, time zones, and simply living together all change who needs to text first. Treat it as a pattern to notice, not a verdict.

The activity heatmap

The heatmap arranges every message into a grid of 7 days by 24 hours, using the timestamp on each message in your local time. Darker cells are the hours when the conversation reliably happens. Most chats have a clear fingerprint: a weekday-evening chat, a weekend-late-night chat, a work-hours chat. Our dedicated guide on activity heatmaps and peak texting hours goes deeper, including how to read weekday versus weekend patterns.

Night owls: messages between midnight and 5 AM

The Night Owls statistic isolates messages sent between 12:00 AM and 5:00 AM. In group chats it instantly identifies the late-night texters; in one-to-one chats it often explains the heatmap better than any other number. Remember that timestamps reflect the sender's activity in the timezone used by the report — if your friend lives abroad, their 2 AM message may have been sent at their lunchtime.

Emojis, reactions, and media

Emoji counts tally the emojis inside text messages, which reveals each person's signature emoji surprisingly fast. Reactions — the hearts and thumbs attached to messages — are stored separately in the export and measure acknowledgement rather than conversation. A person with few messages but hundreds of reactions is present in the chat; they are just participating quietly. Voice notes, photos, reels, and shares appear in the export as their own message types and are not the same as typed text, so a "low message count" friend may actually be the one sending all the videos.

Growth over time

Monthly totals show the shape of a relationship: the intense first months, the steady plateau, the quiet periods, the revivals. Spikes usually have boring explanations — planning a trip, a birthday, a group project — but the long trend is real information. A chat that has grown steadily for two years is a different thing from one that peaked in its first month, and the export lets you see which one you have.

What the statistics cannot tell you

• Deleted and vanished messages that Meta did not include in the export are invisible to every analyzer.
• Intent. A fast reply can mean enthusiasm or an unlocked phone; a slow one can mean indifference or a job. The numbers show behavior, not feelings.
• Anything outside the export. Calls, story replies that never became messages, and chats you did not select are not measured.
• Comparisons across different chats are rough. A group chat and a one-to-one chat have completely different dynamics; compare like with like.

Reading your own report well

Start with the totals, then look at share, then at starters, then at the heatmap. At each step ask "what else could explain this?" before settling on a story. The best use of chat statistics is curiosity — most people finish a report knowing their friendships a little better, not worse. If you have not generated yours yet, the DM export guide walks through getting the file, and the analyzer does the rest in your browser.