A seller asked me a fair question last month: is it really worth staying on top of Instagram messages every evening? Behind the question sat an assumption I hear constantly — that DM buyers are small buyers, so a missed message is a missed trinket.
So we looked at what people actually spend. We pulled the value of orders placed inside DM conversations. One warning first, because it changes how you should read the rest.
Warning: only about 4% of receipts are machine-readable. The other 96% arrive as pictures, and a picture will not tell us its amount. So these numbers are a floor, not a full picture. They show the direction, not the exact size of the market.
What the data says
Here is how order value spreads out across sellers who had priced orders:
| Level | Order value |
|---|---|
| 25th percentile | ~790,000 toman |
| Median | ~1.3 million toman |
| 75th percentile | ~2.3 million toman |
| Top decile | ~4.4 million toman |
| Overall average | ~2 million toman |
Short version: the typical DM basket sits around 1.3 million toman, and the average climbs to 2 million because of the bigger orders pulling it up. Your own currency may differ; the shape of the curve rarely does.
DM is not the impulse-buy aisle
The first thing people get wrong about selling in DMs is picturing tiny, spur-of-the-moment purchases. The data says otherwise. A median of 1.3 million toman means the typical purchase is a considered decision, not a thumb twitch at midnight. And the gap between the median (1.3M) and the average (2M) tells you the distribution leans right: a minority of much larger orders is dragging the mean upward.
That leads somewhere most sellers skip past. An abandoned conversation is not a lost 50,000-toman sale. On the median it is a lost 1.3-million-toman sale. Put that next to what we found about how much revenue burns down over one unanswered question or a late reply, and speed stops being a nice habit. Slow answers are expensive answers.
What actually works
1) Treat every DM lead as a real lead. At a median of ~1.3 million toman, a conversation is worth enough that shaving a few seconds off your reply, or sending one short follow-up, pays for itself. Our customers give you less time than you think — we measured how long DM customers actually wait, and the answer is brutal.
2) Aim at the top decile. The distance between the median (1.3M) and the top decile (4.4M) is room you already have. A matching item, a bundle, or simply showing the better version of what they asked for moves the basket up. Most of that happens in the wording of the reply, which is its own craft — see how a DM sale gets closed.
3) Record the amount, not just the receipt. When only 4% of receipts are readable text, you are flying half-blind. Logging the sale amount as text, instead of parking a screenshot in a chat thread, is what lets you see which product and which conversation actually earns.
4) Show up in person on the big orders. On a multi-million-toman purchase, one human touch at the deciding moment lifts your close rate. Let the AI answer instantly and keep everyone warm; you step in where the basket is heavy.
The takeaway
The DM basket, at least among the orders we can price, sits near 1.3 million toman. That is not pocket change, and it settles the argument: DM is a serious sales channel, not a place where people buy keychains. Since each conversation carries that much value, answering fast and answering correctly is no longer a preference. It is arithmetic.
That is the job Vardast does: a fast, accurate reply to every conversation, so a 1.3-million basket does not evaporate over a few hours of silence. See how Vardast works.
Based on our own analysis of product.fct_business_sales. Only ~4% of receipts are machine-readable text (the rest are images), so these figures are a floor. Order-value distribution measured across sellers with at least 5 priced orders.