How to Connect Klaviyo to Claude for Email and Flow Analytics

Why connect Klaviyo to an AI assistant

Klaviyo holds a remarkable amount of detail about how your email and SMS programme performs. The catch is that getting answers usually means clicking through reports, exporting CSVs, and stitching numbers together by hand. Most marketers only check the dashboards they already know to check, which means underperformers can sit quietly for weeks.

Connecting Klaviyo to an AI assistant like Claude changes the workflow. Instead of navigating menus, you ask a question in plain language and get an answer drawn from your live data. You can interrogate campaign and flow performance, revenue per recipient, list growth, segment health, and deliverability signals without building a single report. More importantly, you can ask the follow-up questions that normally never get asked: why a flow slipped, which segment is quietly shrinking, and what the difference is between your two best-performing campaigns.

What MCP is and why it matters here

An AI assistant on its own cannot see your Klaviyo account. It needs a secure bridge to your data and the actions it is allowed to take. That bridge is the Model Context Protocol (MCP) -- an open standard that lets AI clients connect to external tools and data sources in a structured, permissioned way.

Commerce365 runs an MCP server that exposes your store's data and actions as tools the AI can call on demand. For Klaviyo, that includes tools for campaigns, flows, lists, segments, metrics, and profiles, plus dedicated campaign and flow performance reports. When you ask Claude a question, it calls the relevant tool, pulls the live numbers, and answers from real data rather than guessing. Nothing is dumped into the conversation in advance -- the assistant fetches only what each question needs, which keeps answers current and accurate.

Because MCP is a standard, the same Commerce365 server works with Claude Desktop, the Claude app, ChatGPT, and any other MCP-compatible client. Connect once, use it from wherever you already work.

Step-by-step: connect Klaviyo to Claude with Commerce365

The setup has three stages: install Commerce365, connect Klaviyo, then connect the Commerce365 MCP server to Claude.

1. Install Commerce365 from the Shopify App Store. Find Commerce365 in the Shopify App Store and install it on your store. It is free to install, with paid Plus, Pro, and Max plans available as you grow. Installation links your Shopify store to Commerce365 and creates your account.

2. Connect Klaviyo in the integrations area. Open the Integrations section in Commerce365 and select Klaviyo. Authorise the connection using your Klaviyo API key or OAuth, following the on-screen prompts. Once connected, Commerce365 can read your campaigns, flows, lists, segments, metrics, and profiles. While you are there, you can connect your other channels too -- Shopify is already linked, and Klaviyo sits alongside Meta Ads, Google Ads, Google Analytics 4, and Google Search Console.

3. Connect the Commerce365 MCP server to Claude. In Commerce365, find the MCP connection details for your account. In Claude Desktop or the Claude app, add Commerce365 as a connector using those details and complete the authorisation prompt. Claude will then list the available tools, including the Klaviyo set. From that point on, any question you ask Claude about your email programme can be answered against your live Klaviyo data.

That is the entire setup. There is no spreadsheet plumbing and no custom code -- once the connection is live, you simply ask.

Example prompts to try first

The fastest way to understand the value is to ask real questions. Here are seven prompts that work well once Klaviyo is connected. Use them as written or adapt them to your own date ranges and segments.

• "Which Klaviyo flow drives the most revenue per recipient?" Surfaces your highest-value automation on a per-person basis, not just total revenue, so a small but potent flow is not hidden behind a big one.

• "Compare this month's campaigns to last month -- open rate, click rate, and revenue." A period-over-period read on your broadcast programme without building a comparison report.

• "Which of my segments are shrinking, and by how much?" Catches audience erosion early, before it shows up as a revenue dip.

• "Rank my flows by revenue and flag any whose performance has dropped versus the previous period." Finds the underperformers that deserve a closer look this week.

• "How is my list growing? Show net new subscribers and unsubscribes over the last 90 days." A plain-language view of list health and churn.

• "Are there deliverability warning signs -- rising spam complaints, bounces, or falling open rates -- in my recent sends?" Translates scattered signals into one readable answer.

• "Compare my welcome flow against my abandoned-cart flow and tell me which converts better and why." A like-for-like comparison with a reasoned explanation, not just two numbers side by side.

Because Claude can chain tool calls, you can keep going. Ask it to explain an anomaly, then draft a recommendation, then summarise the whole thing for a colleague -- all in the same conversation.

Let an autonomous agent watch Klaviyo for you

Asking questions on demand is powerful, but you still have to remember to ask. This is where the second half of Commerce365 comes in. Alongside the MCP server, Commerce365 runs autonomous AI agents that monitor your connected data around the clock.

A Klaviyo-focused agent can keep an eye on the metrics that matter -- flow revenue, campaign performance, list growth, segment size, and deliverability -- and surface what changed without you lifting a finger. Instead of discovering a slumping flow at month end, you get a concise weekly digest: which campaigns over- and under-performed, which flows moved, which segments grew or shrank, and where to focus next.

Crucially, agents do not act behind your back. When an agent proposes something -- pausing an underperforming campaign, flagging a deliverability risk, or recommending a segment to re-engage -- it routes that action through an approval queue. You review the proposal, see the reasoning, and approve or dismiss it. You get the vigilance of an always-on analyst with full control over what actually happens.

Getting started

If you run email and CRM on Klaviyo, connecting it to Claude turns a stack of dashboards into a conversation. You ask in plain language, you get answers from live data, and an agent keeps watch in the background so nothing slips. The setup takes a few minutes and the install is free.

Install Commerce365 from the Shopify App Store, connect Klaviyo in your integrations, and link the Commerce365 MCP server to Claude. Then ask your first question -- something like "which flow drives the most revenue per recipient?" -- and see what your data has been trying to tell you.