Design for the Share

The most valuable thing your next graphic can earn is a spot in somebody's messages. Here's the design brief for getting it there.

We all love watching the likes roll in on a fresh post. Here's the news: the algorithm has moved on. Independent analyses estimate a direct message share now carries somewhere between three and five times the weight of a like, and almost nobody designs for it. This post is the evidence and the method.

The three signals that decide your reach

Start with the source. Adam Mosseri, who runs Instagram, has named three signals as the most important for distribution. Watch time. Likes relative to reach. And sends relative to reach, meaning how often something gets shared into a direct message. He's been explicit that for reaching people who don't already follow you, sends matter more than likes. For your existing audience, likes carry a little more weight, and watch time sits at the top for both.

Independent platform breakdowns this year put a number on the gap, estimating a send at three to five times the weight of a like. Treat that range as an estimate rather than a published figure, because platforms don't publish their weights. The direction is well supported. The precise multiplier is fuzzier.

Why a send is worth so much more

A like is nearly free. It costs a tap and it carries no social risk. A send is expensive. You're spending a little bit of your relationship with one specific person, and you're implicitly saying this is worth your attention. That cost is exactly why the platforms weight it: it's a harder signal to fake, and it correlates with genuine value.

So what changes about your design work when the goal is the forward, and the like takes care of itself?

Five design decisions that make a graphic sendable

One. Make one claim per graphic. A graphic that gets sent answers a single question. If your design carries three ideas, the sender has to explain which one they meant, and that friction is where the send dies. Pick the one claim and cut the rest.

Two. Make it legible without the caption. When a graphic arrives in someone's messages, it arrives alone. No caption, no context, no profile. It has to stand up by itself before it can travel by itself.

Three. Design for the small frame. In a message thread the image renders smaller than it does in feed. Test it at about one third the size you designed it. If your smallest text becomes unreadable, make it bigger and cut words until it holds.

Four. Answer a question people are already asking you. This is the highest-leverage decision of the five. People forward answers into conversations that are already open.

Five. Give the recipient somewhere to go. A sendable graphic gets seen by people who've never heard of you. Put your handle, small, in the same corner every time. One legible identifier. A watermark across the middle reads as defensive, so skip it.

Then run one test before you publish. Would somebody send this to a friend without typing an explanation first? If they'd need to add context, the graphic needs another pass. Go back to decision one.

One claim, legible alone, sized for the thread, answering a live question, signed in the corner. Make one idea complete inside a single frame, and the sending takes care of itself.

Design for the person who receives it, and the person who sends it will find you.

Put this into practice

• Audit your last five graphics: how many carry exactly one claim? Rework any that carry two or more.
• Run the caption-hide test on your next design before you publish it.
• View your design at one third its size and confirm the smallest text still reads.
• Collect the three questions you've been asked most this month and turn the top one into a single-frame answer.
• Pick one corner for your handle and place it identically on every graphic going forward.
• Before every publish, ask: would somebody send this without typing an explanation first?