AI is everywhere.
Our data shows there is still resistance.

The gap between narrative and data

Every vendor deck, every conference keynote, and every earnings call suggests that AI in ecommerce support is mainstream.

In our 2026 State of Conversational Commerce survey, the numbers back that up: 96% of ecommerce brands named customer support automation as their top AI use case. 93% have used AI for at least a year. 83% say it is already transforming their business or bringing steady improvements. Zero percent said they were still hesitant.

Then we measured what is actually happening. Using platform-level data (not self-reported usage or survey responses) we looked at whether AI is actively deployed and resolving customer interactions across ecommerce brands.

The answer: roughly 1 in 5. Despite near-universal stated adoption, most ecommerce brands have not yet made the organizational decision to deploy AI in customer-facing support.

At the same time, the adoption curve shows steady month-over-month growth, increasing from 12.3% in April 2025 to 17.6% by January 2026.

Who has adopted, and who hasn't

Brand size is the strongest predictor of AI adoption. The bigger the brand, the more likely they have deployed AI in support. The relationship is nearly linear across every GMV tier.

Fewer than 18% of brands under $1M GMV have deployed AI. That number crosses 25% around $5M GMV, 35% at $25M, and approaches 50% for brands above $500M GMV. Even at the largest tier, half have not deployed.

The data suggests this is less about technology readiness and more about organizational capacity. Larger brands are more likely to have dedicated CX operations teams, implementation resources, and the internal mandate to run a structured rollout. The gap is not about whether AI works. It is about whether a brand has the setup in place to deploy it properly.

But the brands that have made that leap, regardless of size, are the ones showing up in the performance data that follows.

What adopters are seeing

Among brands that have deployed AI, the response time improvement is not incremental. It is structural.

Brands automating near 0% have a median first response time of 736 minutes. At 20% automation, that drops to 400, roughly half. At 30%, it falls to 80 minutes. At 40%, 12 minutes. The gains do not scale evenly. They accelerate.

The results are not just incremental. First responses didn't only get faster -entire batches of tickets essentially became one-touch because AI could handle and clear them from the queue.

The headcount equation

Faster response times are visible to customers. The efficiency gains are visible to finance.

We calculated "AI Agent Equivalents" by measuring how many automated tickets AI handles relative to the average human agent's workload. The result: AI is not just handling tickets -it is doing the equivalent work of multiple human agents.

At 50%+ automation, brands operate with just 3 team members doing work that would typically require 9+. AI does the equivalent of 6.3 agents. The human team stays the same size but shifts to higher-value work.

The bottom line

Most brands are not there yet. Those that have committed are responding 10x faster and scaling without growing their team. The ones still hesitating are hiring their way through the same volume.

The performance gains in this report are not exclusive to enterprise. They follow from deployment, not scale. Brands that invest in proper setup (clear SOPs, structured knowledge bases, and a deliberate rollout) can reach the same automation tiers regardless of team size or GMV. The barrier is organizational commitment, not budget.