How Are AI Agents Changing Ecommerce Agency Retainers?

AI agents are moving ecommerce agency retainers from billable hours to outcomes. When agents find problems and draft changes across a client store, each unit of work costs less, so selling time stops making sense. Retainers are shifting toward a monthly scope of shipped, measured improvements, with the agency as reviewer and merger of record.

Ecommerce agency retainers have long been sold as time: a set number of developer and strategist hours each month. AI agents change the math behind that model. This guide explains why the hours retainer is under pressure, what agencies are replacing it with, and what merchants should agree with their agency. It is part of a series on the AI-native ecommerce agency, an agency that runs AI agents as core infrastructure while its people review and own the work.

Why are AI agents putting pressure on billable-hours retainers?

A billable-hours retainer prices the effort of finding, reproducing and building each change. AI agents now do much of that effort. They read session, error, speed and conversion data from a store, surface what is broken or underperforming, and draft the change. When that happens, the hours an agency would have billed for investigation shrink, and a client paying for hours starts asking what the hours are for.

John Duncan, CEO of the Salesforce Commerce Cloud agency 64labs, put the question bluntly in Managed Agents, Not Managed Services (August 31, 2026). He writes that a retailer watching agents do the work will ask why it is paying a body-shop margin on bodies, and that "the billable-hours model doesn't survive that question." In the same article, he estimates the resulting cost reduction at 50 to 80 percent, an estimate from his own agency's experience rather than an industry study.

Clients are already adjusting their expectations. Rick Watson, founder of RMW Commerce, said on BigCommerce's Keeping Commerce Weird podcast episode The Agency Survival Playbook in the Age of AI that "what AI is doing is reducing everyone's expectations of what they should pay." For an agency selling hours, that is a pricing problem before it is a capability problem.

What is replacing the hours retainer?

Agencies are moving ongoing work toward a retainer priced on scope and results. Instead of a bank of hours, the client pays for a monthly scope of improvements: agents find and draft the changes, the agency reviews, merges and ships them, and the monthly report shows what each change moved. Hours still have a place for discovery and large builds, where effort is hard to predict.

Most agencies will run more than one model at once. Watson described the shift as "more of a pricing model difference than a capabilities difference," and noted that pure outcome pricing is often limited by the client, not the agency. A performance share needs clean measurement and an agreed baseline, which many merchants do not have yet. The outcome-scoped retainer sits in the middle: predictable for the client, and tied to shipped work rather than time.

Noibu sees the same direction in its conversations with agencies. One commerce agency told Noibu in November 2026 that it is moving from selling hours to selling business value. In another case, a merchant that had paid its agency as needed for years put the agency on a monthly retainer during a Noibu trial, once there was a steady list of agent-drafted changes to review and ship.

Do AI agents make agency retainers smaller?

Not necessarily. Lower cost per change usually means more changes, not a smaller relationship. Duncan argues in the same 64labs article that "the cost per unit of work is collapsing, so the volume of commissioned work goes up, not down." A merchant that once funded a handful of tickets a month can now fund a steady flow of improvements across checkout, speed, product pages and experiments.

Agencies that package that flow well can grow the retainer. Blue Badger, a Montreal ecommerce agency, made Noibu part of its retainer offering and used it in regular client reviews to show issues it had found and resolved; the agency reports a 20% increase in client retainer value. The full story is in how Blue Badger increased client retainers by 20% with Noibu.

The work itself also changes shape. When agents carry the repetitive finding and drafting, the agency's people spend more time on review, prioritization and strategy. Duncan's article states the rule plainly: "A named human reviews every merge." That review role is what the client keeps paying for.

Why do outcome-based retainers go wrong?

Outcome-based retainers fail for predictable reasons, and most are agreed before any work starts. The common ones:

• No baseline. If nobody measured conversion or speed before a change, nobody can agree what it moved afterward.
• Unclear approval. If the merchant, the agency and the agents all touch the store without a named reviewer, changes stall or ship without a clear owner.
• Billing that only rewards requests. An agency paid only for work the client asks for has little reason to run agents on its own schedule, so problems wait until someone files a ticket.
• Reporting hours instead of results. A monthly hours log tells the client nothing about what changed for shoppers.

Problems that wait for a ticket are expensive because shoppers rarely report them. Over 90% of website errors are never reported, so a retainer built on client requests misses most of what is going wrong on a store.

What should an AI-era agency retainer include?

An AI-era retainer is easiest to manage when five things are written down. This list works for merchants reviewing a renewal and for agencies redesigning their offer:

1. Scope: the areas the retainer covers each month, such as checkout errors, page speed, product pages and experiments.
2. Cadence: how many improvements the agency aims to ship per month, ranked by revenue at stake.
3. Approval: who reviews and merges each change, and how the merchant signs off on what is in scope.
4. Measurement: the before and after metric for each change, measured on real shopper traffic.
5. Reporting: a monthly list of what shipped and what it changed, in place of an hours log.

For agencies that control a client's repository, see how to connect AI agents to GitHub when you do not own the repo and how to keep AI-generated pull requests reviewable. For the measurement step, see how to know if a website change actually improved conversion.

What should a merchant ask its agency about AI and the retainer?

A merchant does not need a new agency to benefit from AI agents. It needs a conversation with its current one. Useful questions: Which parts of the monthly work do agents find and draft? Who reviews and merges each change? What will the monthly report show besides hours? How will each change be measured? Which work still makes sense to bill by the hour?

An agency that can answer those questions clearly is usually already moving toward an outcome-scoped retainer. An agency that cannot answer them yet may simply need the tooling and data to get there, and the merchant can help by giving it access to the same store data.

How does Noibu fit into an agency retainer?

Noibu's six supervised agents do the finding and drafting that an outcome-scoped retainer depends on. The CRO Agent finds underperforming product pages and rewrites them, with inline edits that publish to Shopify in one click after approval. The Bug Resolution Agent finds the top errors shoppers hit, traces root causes in the codebase and opens a pull request for review. The Performance Agent measures Core Web Vitals from real shopper traffic and stages changes, the ADA Agent audits high-traffic pages against WCAG, the A/B Testing Agent builds and runs experiments and reports the winning variant, and the ROAS Agent scores paid traffic on orders rather than clicks.

Every Noibu agent runs the same loop: find, build, approve, ship, measure. Nothing ships without approval, so the agency stays the reviewer and merger of record, and the measure step gives the agency the before and after results its monthly report needs. The agents work in the Noibu console and in AI tools teams already use, such as Claude, ChatGPT, Microsoft Copilot, GitHub Copilot and OpenAI Codex.

Noibu is AI for ecommerce operations, the always-on optimization system that keeps online stores running at their best. A team of supervised AI agents finds what's broken, slow, or leaking revenue, does the work, and hands results back for human approval. Trusted by retailers like Free People, Guess, Liquid I.V., and Ariat.

Frequently asked questions

Related reading

• What is an AI-native ecommerce agency?
• How Blue Badger increased client retainers by 20% with Noibu
• CRO agency, in-house team, or AI agents? How to split the budget

See what agents would find on your store: run a free website audit →