Email isn't dying, chat is doubling. But it's not that simple.

Email held flat. Chat ate Social DM's lunch.

You have read the LinkedIn posts. A generation of shoppers thinks in conversations, not tickets. WhatsApp is the new inbox. Email is for boomers.

We pulled the data on every customer support ticket on Gorgias across the last 24 months. Then we held the brand cohort steady so growth at the platform level could not flatter or mask the channel mix. Same accounts. Same 4-month window two years apart. Ticket volume and share per channel.

‍

The shift is real but partial. Email is structurally sticky. It absorbs every customer who has an email address and does not want to start a chat session. That is most customers, most of the time.

Email volume held roughly flat over two years. The conversational stack grew on top, but most of chat's growth came at the expense of Social DMs, not email. The channels customers used to send a casual question through (Instagram DM, Facebook Messenger) are giving way to chat on the brand's own site. Email keeps doing what it has always done.

Share figures exclude Social public, where a Sep 2025 ingestion change distorts the denominator. Caveat detailed in methodology.

The median brand is still 85% email.

Headlines aggregate. Brands do not. Across active brands on the platform, the distribution of channel mix is heavily skewed.

‍

Half of active brands have less than 2% of their support volume on chat. Ten percent of brands do all of their support through email. The brands that have made chat 40%+ of their volume are roughly 5.5% of the active base.

The channel shift is a leadership move. Not a mass migration.

Where AI resolves, brands concentrate.

The most useful question is not "is chat growing." It is "why are some brands betting on chat." Look at where AI actually handles tickets.

‍

One in four chat tickets are resolved by AI. One in twenty-six email tickets are. AI handles chat 6.7x more often than it handles email.

Chat as a channel is structurally compatible with automation. Sessions are short. Intent is narrow. The shopper expects an instant response, which removes the awkwardness of an AI reply. Email tickets are longer, more layered, more emotionally charged, and arrive with the implicit expectation of a considered human response.

The brands concentrating chat are not chasing a trend. They are stacking volume into the channel where AI does the work. And the share of brands doing it rises with GMV.

‍

A quarter of brands above $30M GMV have made chat a meaningful piece of their channel mix. Fewer than one in eight under $5M have. The pattern tracks the AI adoption curve from Report 1: the brands with the operations capacity to deploy AI Agent properly are the same brands with the operations capacity to commit to chat as a deliberate channel. The two moves cluster.

The performance gaps are real but not the headline.

Chat is fast. Email is slow. CSAT is closer than the speed gap suggests.

‍

Chat is 38x faster than email on first response, 8x faster on resolution. CSAT on chat is 2.5 points lower than email. Speed and satisfaction do not move together the way the narrative suggests. Chat customers are faster to ask, faster to be answered, and faster to score a brand harshly.

This does not undermine chat. It clarifies the trade. A brand investing in chat is investing in throughput and AI leverage, not in satisfaction lift. The CSAT case for chat is neutral. The economics case is decisive.

The bottom line

The channel shift is real. The shape of it is not the one the narrative predicts.

Email volume held flat on the brands in this analysis. Chat grew on top of it, taking volume from Social DMs, not from the inbox. The spread between email and the conversational stack is widening because the conversational stack is rebalancing internally, not because email is dying.

Your decision is not whether to abandon email. It is whether to commit to chat at a scale that earns the leverage.

Methodology

Platform-level behavioral data from Gorgias merchants. Customer-status accounts only. Spam excluded except where AI Agent billed. Channel of first ticket message. Channels grouped: Email (email, contact_form), Chat (chat, custom-chat), Voice (phone, aircall), SMS, WhatsApp, Social DM (Instagram DM, Facebook Messenger, Twitter DM), Social public (Instagram comments, Facebook public, mentions, ad-comments, TikTok Shop). Matched-cohort YoY uses accounts with at least 60 tickets across both 2024 Jan-Apr and 2026 Jan-Apr windows (10,990 brands). Per-account distribution and per-channel performance use the last 90 days minus 7-day billing lag, minimum 30 tickets per account. AI Agent fields clamped to ticket_created_date >= 2025-04-25 per the AI Agent architecture cutoff. FRT and resolution time are medians per account, then medians across accounts; non-positive durations excluded. CSAT positive rate is score >= 4 over surveys returned; response rates on Voice and WhatsApp are low, biasing those CSAT figures upward. AI resolution rate is is_ai_agent_billed over total tickets in channel. Social public ingestion changed in late 2025 (volume per stable account base dropped 4x in Sep 2025). Volume comparisons are absolute matched-cohort counts. Share figures are computed on the matched cohort with the Social public channel excluded from the denominator, because including it would credit email and chat with apparent share gains driven by an upstream ingestion artifact rather than customer behavior. Caveat: brands deploying chat at high volume tend to be brands with the operational capacity to deploy AI Agent; the AI-resolution-by-channel statistic is descriptive of where AI is deployed today, not a claim about AI capability by channel.