Why Claude + Klaviyo + Shopify is the ecommerce stack to watch
If you run a Shopify store, you already know the truth about email: it's still the highest-ROI channel you own, and Klaviyo is where most of that revenue gets made. The catch is that good Klaviyo work is slow. Segments, flows, campaign copy, the back-and-forth on the offer. It all eats the week. So most stores ship a welcome flow, an abandoned-cart flow, and a handful of campaigns, then never touch them again.
Claude changes the economics. Connected to your store, it can read what's actually selling, reason about who to talk to and why, and draft the whole thing (segments, flows, and campaign emails) in the time it used to take to write one subject line. The only thing it shouldn't do is press send on its own. That's the job this workflow is built around: Claude does the heavy lifting; you approve; Klaviyo ships.
The missing piece is a place for all of that to live. Drafts inside a chat window scroll away, and an agent firing changes straight into Klaviyo with no review is a recipe for a 2am "why did we email everyone twice" moment. Plainpaper is the canvas in the middle: every segment, flow, and email the agent produces lands as a typed card you can see, question, and approve, with the Shopify data and the offer that shaped it attached.
What each tool does in this workflow
Each tool has a clear lane, and the workflow is strongest when you let them stay in it.
The workflow, end to end
Six stages, each producing a persistent artifact that feeds the next. The whole thing runs from the chat window you already use. Claude does the work and writes cards onto the board in real time.
Claude · Q3 winback board
shopify · read_checkouts
klaviyo · create_flow
Drafted a 3-email flow and put it on the board for review.
What you'd see: Claude calls Shopify and Klaviyo over MCP, then drops the drafted flow onto the board. Nothing is live yet.
Step by step: a campaign from brief to send
Connect Claude to Shopify, Klaviyo, and a board
Everything here runs over MCP, the open protocol that lets Claude call other tools. In your Claude client, connect three servers: your Shopify store (read), Klaviyo (read + write), and a Plainpaper board. Now the same conversation can look up a product, draft an email, and put it somewhere you can approve it.
Permissions matter. Give Shopify read-only access and keep Klaviyo's send/publish actions behind your approval on the board. The agent can draft a campaign freely; it can't ship one without you.
Let Claude read the store before it writes a word
Generic AI copy reads generic because the model is guessing. Ground it first. Ask Claude to pull your bestsellers, slow movers, margins, repeat purchase rate, and the products that drive first orders. That single step is the difference between "Shop our new collection" and copy that leads with the exact product a segment actually buys.
Read my Shopify store and summarise on the board: top 10 products by revenue (last 90 days), the 5 best first-order products, repeat-purchase rate, and any SKUs with stock to clear. Flag what you'd build a campaign around and why.
Build the Klaviyo segments
Ask Claude to translate what it found into segments and create them in Klaviyo. Good starting set: VIPs (top 10% by lifetime value), one-time buyers, lapsed (no order in 90+ days), category browsers, and engaged non-purchasers. Each one becomes an Audience card on the board, so you can see the logic and the size before anything is built on top of it.
Create these Klaviyo segments and add an Audience card for each: VIPs by LTV, one-time buyers, lapsed 90d, browsed-but-didn't-buy. For each, note the hook, the offer you'd use, and what we should never say to them.
Draft the flows and campaigns
Now the writing. Have Claude draft your core flows (welcome, abandoned cart, browse abandonment, post-purchase, win-back) plus any seasonal campaign. Each email arrives as a card with subject-line variants, preview text, body, and CTA, linked back to the segment and product it targets. Because the lineage is visible, you can tell at a glance whether the offer fits the audience.
Draft the abandoned-cart flow (3 emails: 1h, 24h, 48h). Email 2 should surface the specific items left in the cart from Shopify. Give me two subject lines per email and keep the tone warm, not pushy. Put each as an Email card linked to the flow.
Approve on the board: the safety layer
This is the step that makes an autonomous agent safe to run on real revenue. Every email sits on the canvas as a finished, uploadable asset with its full lineage: the audience, the insight, the product, the offer. Read it, tweak the copy in plain language, then approve. Only approved cards get scheduled.
The agent can draft a hundred emails in a minute. The board is where one human decides which of them are good enough to send.
Pull results back and let the next campaign compound
After a send, ask Claude to pull Klaviyo opens, clicks, and Klaviyo- attributed Shopify revenue back onto the board as Result cards. Now the next campaign starts from "the win-back at 15% off beat 20% off on margin, do that again" instead of a blank chat. Your learnings stop evaporating.
The five flows to build first
If you do nothing else, build these. They have the clearest triggers, the best Shopify data to draw on, and the fastest payback, which makes them the perfect first job for the agent.
- Welcome flow: fires on signup. Claude leads with your best first-order product and your brand's actual story, not a generic "10% off."
- Abandoned cart: fires on checkout-started. Email 2 names the exact items from the Shopify cart; the recovery rate lives or dies on that detail.
- Browse abandonment: fires on product-viewed. Softer than cart; Claude matches the email to the category browsed and suggests two alternatives.
- Post-purchase: fires on order-placed. Confirmation, then education, then a cross-sell timed to your product's usage cycle.
- Win-back: fires for lapsed buyers. Claude reasons about the right incentive against margin instead of defaulting to the biggest discount.
How this compares to Klaviyo's built-in AI and a plain chatbot
Klaviyo's own AI is good at what it does. So is asking ChatGPT for a subject line. Neither replaces a stack that reads your store, reasons across the whole campaign, and keeps a human in the approval loop.
| Approach | What it's good at | Where it falls short |
|---|---|---|
| Plain chatbot copy/paste into Klaviyo | Fast first drafts of copy and ideas. | No store context, no memory between chats, every campaign starts from zero. |
| Klaviyo built-in AI | Subject lines, send-time and predictive optimisation inside Klaviyo. | Optimises within Klaviyo; doesn't plan a campaign or reason across your offer and margin. |
| Claude + Klaviyo + Shopify + Plainpaper | Reads Shopify, plans and writes the whole campaign, keeps lineage, approve before send. | You still set the strategy and approve. By design, not by limitation. |
Pitfalls to avoid
- Letting the agent send. Keep publish actions behind board approval. An agent that drafts is an asset; an agent that ships unreviewed is a liability.
- Skipping the Shopify read. If Claude writes before it reads the store, you get generic copy. The data step is the whole point.
- Discounting on reflex. Ask Claude to weigh incentives against margin. "Bigger discount" is rarely the right answer for VIPs.
- One giant prompt. Work stage by stage (segment, then flow, then copy) so each artifact is reviewable instead of a wall of text.
Key takeaways
- Connect Claude to Shopify (read), Klaviyo (read+write), and a Plainpaper board over MCP.
- Always let Claude read the store before it writes. Grounded copy beats generic copy every time.
- Build segments and the five core flows first; they recover the most revenue fastest.
- Approve every email on the board. The agent drafts; you decide what ships.
- Pull results back so each campaign compounds on the last instead of starting from a blank chat.
Frequently asked questions
Can Claude build Klaviyo flows and campaigns?
Yes. Connected over MCP, Claude can read your Shopify store and draft Klaviyo segments, flows, and campaign emails: subject lines, preview text, body, and CTAs. It produces finished, approvable drafts; a human approves them on the board before anything is scheduled or sent in Klaviyo.
Does Claude read my Shopify data?
With a Shopify MCP connection, Claude can read catalog, pricing, orders, and customer data so its segments and copy reference real products and real buying behaviour instead of generic placeholders. It only takes write actions you approve.
Is this better than Klaviyo's built-in AI?
They solve different problems. Klaviyo's AI optimises inside Klaviyo (subject lines, send times, predictive segments). Claude reasons across the whole campaign: it reads Shopify, weighs the offer and margin, plans the sequence, and writes the copy. Plainpaper keeps that reasoning visible and approvable instead of trapped in a chat window. Use both.
Which Klaviyo flows should I build with Claude first?
Start with the flows that recover revenue: abandoned cart, browse abandonment, welcome, and post-purchase. They have the clearest triggers and the fastest payback. Once those run, add win-back and a replenishment flow tuned to your product's reorder cycle.
Will Claude send emails without my approval?
No. The workflow is built so the agent only ever drafts. Every email lands on the board as a finished asset with its lineage, and you approve it there. Only then does Claude schedule or push it through Klaviyo.
Do I need to know how to write prompts?
Not really. You brief Claude the way you'd brief a marketer (the goal, the offer, the audience) and steer from there. The board shows you exactly what it produced, so you correct in plain language instead of re-engineering a prompt.
Next step: give the workflow a home
You have the workflow. The piece that makes it safe and repeatable is the board in the middle, where the agent's work becomes visible, linked, and approvable. Plainpaper is free to try and connects to your Claude client over MCP in a couple of minutes. Your next Klaviyo campaign can start from last campaign's revenue instead of a blank chat.