- eCommerce marketing automation pays only when every tool reads from the same customer record: orders, email and SMS engagement, consent and ad audiences tied to one profile per person.
- Map each connection in your stack by level (native, official app, browser only, middleware, API only or gap), then fix the ones that carry orders, refunds and consent first.
- Send the core events (product views, carts, checkouts, orders, refunds) from the store to your email and SMS platform, server-side where possible, and push segments from that profile back out to Meta and Google.
Why do disconnected marketing tools leak profit?
Marketing automation means messages and actions that fire on their own when a customer does something, or stops doing something. When your store, email platform, SMS tool and ad accounts each keep their own version of a customer, automation starts making expensive mistakes.
The leaks are usually small and constant:
- A customer buys, but the order reaches the SMS tool late. They still get the “15% off, your cart is waiting” text and use the code on their next order.
- Meta keeps retargeting people who bought yesterday, because the purchase audience refreshes slowly or not at all.
- A repeat customer receives the first-order welcome discount because the signup form doesn’t know they’ve bought before.
- Refunds never reach the email platform, so your “VIP” segment includes people who return most of what they buy.
- Nobody can calculate customer lifetime value with confidence, because orders, messages and ad spend live in separate places.
What does this cost? An illustrative example
Take a store with 2,000 orders a month and an average order value of $80. These numbers are for illustration only.
| Leak | Assumption | Monthly cost |
|---|---|---|
| Cart discount sent to people who already bought | 5% of orders (100) use a 15% code they didn’t need | $1,200 in margin |
| Retargeting spend reaching recent buyers | 20% of a $3,000 retargeting budget | $600 in wasted spend |
| Welcome discount used by repeat customers | 40 orders with a 10% code | $320 in margin |
| Total | $2,120 a month, about $25,000 a year |
None of these show up as a line item, which is why they persist.
What does “one customer profile” actually mean?
One customer profile is a single record per person, keyed on email address and phone number, that every tool reads from or writes to. At minimum it holds:
- Identity: email, phone, store customer ID, name, location.
- Consent: email and SMS consent stored separately, with source and timestamp.
- Orders: count, revenue, first and last order date, products, discounts used, refunds and returns.
- Behavior: product views, add to carts, checkouts started, site visits.
- Engagement: email and SMS opens, clicks, unsubscribes, bounces.
For most Shopify stores between $1M and $25M in revenue, the email service provider (ESP) acts as the working profile. Klaviyo, for example, merges Shopify orders and on-site events into one profile and lets you segment on all of it. Shopify stays the source of truth for orders and customer records. A dedicated customer data platform starts to make sense when you run several storefronts, sell through retail or wholesale, or have more tools than the ESP can connect cleanly.
What a connected profile lets you do
The payoff is segmentation that mixes value and intent. Our Value × Intent (RFMB) model combines classic RFM analysis (recency, frequency, monetary value) with a Buying Stage layer: B1 visit, B2 product view, B3 add to cart, B4 checkout, B5 purchase, all within the last 30 days.
With that in one place, a high-value customer at B4 gets a personal checkout reminder with no discount. A first-time visitor at B3 with a large cart might get a free-shipping nudge. Neither works if orders and on-site events live apart.
What are the integration levels between eCommerce tools?
Before you fix anything, label each connection by how it actually works. An API integration is any connection where two systems exchange data directly through their programming interfaces; the levels below describe who builds and maintains it.
| Level | What it means | Typical example | Reliability | What to watch |
|---|---|---|---|---|
| Native | Built and maintained by the platform vendors, server to server | Shopify and Klaviyo, Shopify and Shopify Email | High | Which events and fields actually sync |
| Official app | Vendor-built app or plugin you install | Meta’s Facebook & Instagram app on Shopify, Klaviyo for WooCommerce | High to medium | Permissions, version updates, which events are server-side |
| Browser only | A pixel or script in the theme, nothing server-side | An ad pixel pasted into theme code | Medium to low | Ad blockers, consent banners, browser limits, no refunds |
| Middleware | Zapier, Make or similar passes data between tools | Subscription app events sent to the ESP | Medium | Task limits, silent failures, field mapping drift |
| API only | The tool has an API but no ready connector | Custom ERP, niche loyalty tool | Depends on the build | Needs a developer, monitoring and documentation |
| Gap | No connection at all; CSV exports or nothing | Wholesale orders, a separate retail POS, helpdesk tickets | None | Data is stale or missing |
If a purchase is recorded only by a script in the customer’s browser, any blocker, consent refusal or closed tab means the event never arrives. Server-side tagging and server-to-server integrations avoid most of this.
How do you map your marketing stack?
Work through it in this order, in a spreadsheet:
- List every tool that touches customer data. Store platform, ESP, SMS tool, reviews, loyalty, subscriptions, helpdesk, ad platforms, analytics and your consent banner.
- Record each connection. For every pair of tools that should share data: direction, integration level, which events and fields move, and how fast (real time, hourly, daily).
- Note the identity key. Does the tool match people on email, phone or store customer ID? Mismatched keys create duplicate profiles.
- Trace consent. Where is SMS consent captured, and does it reach the SMS tool and the store’s customer record?
- Test with a real order. Place a test order with a new email, refund it, and check where each step appears and how fast.
Here is what a partial map might look like for a typical Shopify store (illustrative):
| Connection | Level | Events | Latency | Status |
|---|---|---|---|---|
| Shopify to Klaviyo | Native | Orders, checkouts, refunds, product views | Near real time | OK |
| Shopify to Meta | Official app | Purchase, add to cart, checkout | Near real time | Check data-sharing setting |
| Reviews app to Klaviyo | Official app | Review submitted | Near real time | OK |
| Subscription app to Klaviyo | Middleware | Subscription cancelled | Every 15 minutes | Fails silently on errors |
| Wholesale orders to Klaviyo | Gap | None | None | B2B buyers get DTC promos |
Which events should flow to which tools?
Event tracking is the record of what each customer does. Send each tool the events it needs to make a decision, not everything.
| Event | Source | Send to | Why it matters |
|---|---|---|---|
| Viewed Product | Store (on-site tracking) | ESP | Browse abandonment, Buying Stage B2 |
| Added to Cart | Store (on-site tracking) | ESP | Cart abandonment, B3 |
| Started Checkout | Store (server) | ESP, ad platforms | Checkout abandonment, B4 |
| Placed Order | Store (server) | ESP, SMS tool, Meta, Google Ads, GA4 | Stops abandonment flows, starts post-purchase, feeds ad bidding |
| Fulfilled or Delivered | Store or shipping app | ESP | Times review requests and cross-sells |
| Refunded or Cancelled | Store (server) | ESP, ad platforms, GA4 | Corrects revenue and customer value |
| Subscribed with consent | Signup form or checkout | ESP, SMS tool, store customer record | Keeps one consent record |
Placed Order is the most important event in the list. Every tool that sends messages or buys ads should know about a purchase within minutes. Feed it into GA4 and your ad platforms server-side so bidding learns from real orders.
From the profile back out to ads
The connection should also run the other way. Most ESPs, Klaviyo included, can sync segments to Meta custom audiences, and many also support Google Customer Match:
- Exclude first. Remove recent buyers and active subscribers from prospecting and retargeting campaigns.
- Build seed audiences. Use your highest-value customers as the source for a lookalike audience, not your full list.
- Respect consent. Only sync people whose consent and your privacy policy cover that use.
How does setup differ on Shopify, WooCommerce and Magento?
Shopify. Install your ESP from the Shopify App Store and enable its app embed so on-site events (product views, add to carts) are tracked. Confirm SMS consent collected at checkout syncs to the SMS tool. If your ESP sends abandoned checkout emails, turn off Shopify’s own abandoned checkout notification to avoid duplicates. Use Shopify Flow for internal rules such as tagging customers or alerting the team. Connect Meta and Google through their official sales channel apps and check the data-sharing settings.
WooCommerce. Most connections are plugins. Test that checkout events fire on both the classic and block checkout if you use either, and that caching or security plugins don’t block on-site scripts. Expect to rely more on middleware for secondary tools.
Magento / Adobe Commerce. Connections usually come as extensions, and syncs often run on cron jobs, so check the schedule. Plan developer time for API-only tools, and retest every integration after a platform upgrade.
What should you fix first?
Rank fixes by profit impact. A practical order:
- [ ] Placed Order and Refunded events reach every tool that sends messages or buys ads, within minutes.
- [ ] SMS and email consent sync to one record, in both directions.
- [ ] Started Checkout and Added to Cart arrive in the ESP, server-side where the integration allows.
- [ ] Recent buyers are excluded from ad audiences automatically.
- [ ] Secondary tools (reviews, loyalty, subscriptions, helpdesk) connect through official apps or monitored middleware.
- [ ] Gaps such as wholesale or retail orders get connected if they represent meaningful revenue.
If you’d rather not map it by hand, the Free Profit Audit scans your store, shows which tools in your stack don’t talk to each other and estimates what the gaps cost per year. When you’re ready to build the connected setup and the flows on top of it, our Lifecycle Retention team does that work.
FAQ
Do I need a customer data platform to get one customer profile?
Usually not at first. For a single Shopify store, an ESP like Klaviyo with native Shopify sync can hold the working profile. A CDP becomes worth it when you have several storefronts, offline sales or more tools than the ESP connects well.
Is Zapier or Make good enough for connecting eCommerce tools?
For low-volume, non-critical events, yes. For orders, refunds and consent, prefer native or official integrations, because middleware can fail without alerting anyone. If you use it, assign an owner and set up error notifications.
Which tool should be the source of truth for customer data?
Your store platform should own orders, refunds and the customer record. Your ESP should own engagement and act as the working profile for segmentation. Write down which system wins when two disagree, and keep it consistent.