Field notes from the founders
Klaviyo attributed revenue vs Shopify: why email claims more than you sold
November 15, 2026
Every year-end review has a version of this meeting. Growth reports CPA down 30%. Retention has moved 30% more of the base onto subscription and launched a loyalty scheme. Supply chain has taken $500,000 of cost out, and finance has improved cash flow by 20%. It all sounds great. Then the year closes with revenue up 3%, profit flat and EBITDA down 10%. Every team measured something real, each measured something different, and none of it adds up.Your marketing dashboards hold the same meeting every month. Add up the revenue Meta, Google and Klaviyo each say they drove and you'll often pass 100% of what Shopify took. Klaviyo is frequently the biggest claimant, which is why Klaviyo attributed revenue vs Shopify actual revenue is one of the most common questions on the Shopify and Klaviyo forums, usually asked by someone whose boss has just noticed that email "drove" 45% of sales.Attribution has no right answer. No model can say for certain which ad, email or search deserves the credit for an order, and anyone promising otherwise is selling a model with a confident name.Platforms also disagree for a second, under-appreciated reason: they collect different data before any credit is handed out. Browser pixels against backend orders, gross against net, refunds booked against the original order or on the day they're processed, cookie consent deciding who each tool can see, first-party tags against third-party ones that Safari and iOS treat less kindly. Klaviyo has both problems. Its rules for handing out credit are generous, and the revenue it counts differs from Shopify's.Klaviyo uses last-touch attribution by default: all of the credit for an order goes to the last eligible message interaction before it. For new accounts, Klaviyo's help centre lists a five-day window for an email open or click, five days for an SMS click and 12 hours for an SMS delivery (checked November 2026). You can change them, up to three times in 30 days, and Klaviyo recalculates your history when you do.An email open counts as a touch, and so does a text message that was merely delivered. Neither needs a click, and neither needs the customer to reach your site through Klaviyo. Klaviyo matches the purchase to the customer's profile, so someone who opens your newsletter on their phone on Monday, then searches your brand on their laptop on Thursday and buys, is an email conversion. Google Ads, which charged you for the brand search click, counts the same order as its own. That's a generous definition of influence.Apple opens your emails for you. Apple Mail Privacy Protection, which arrived with iOS 15 in 2021, downloads email content in the background, tracking pixel included, whether or not the email is read. It applies to anyone reading in Apple's own Mail app, whichever provider the account is with (a Gmail address read in Apple Mail is covered, while the Gmail app on an iPhone isn't). Users are asked to switch it on when they first open Mail, and very few say no. Litmus put Apple at 76% of all email opens in October 2026 and estimates that privacy-protected opens make up roughly 55% to 60% of opens overall. Brevo's 2025 benchmarks put the ecommerce open rate at 15.5%, or 30% once Apple's automatic opens are included, so about half of reported ecommerce opens never involved a human. Klaviyo counts these opens towards attribution unless you exclude them.Campaigns and flows are timed for when people buy. Abandoned checkout, browse abandonment, replenishment reminders and win-backs are built to land just before a likely purchase. So are the campaign calendar's big moments: Black Friday, the festive run-up, payday and month-end pushes. That's good marketing, and it's also where last-touch attribution hands out the most credit. Some of those customers needed the nudge. Many were buying anyway, and with an open or a delivery counting as a touch, the email barely had to be noticed to collect the order.Retention brands email their best customers often. If a loyal customer gets three campaigns a week, there's nearly always a send within five days of anything they buy. For a high-retention brand, where repeat customers bring in most of the revenue, that puts most repeat orders inside an open email window.Subscription renewals get claimed too. Klaviyo's Placed Order metric includes subscription orders from Recharge or Skio. A renewal that was always going to charge on schedule gets credited to whichever email the subscriber opened that week, including the reminder that their next box ships on Friday. Crediting email with a renewal it had nothing to do with is shooting fish in a barrel, and on a subscription brand it can be a big share of the total. We'd strip renewals out of email's revenue entirely (how, below).Transactional messages count unless you say otherwise. Klaviyo lets you exclude interactions with transactional messages, such as order confirmations and shipping updates, from attribution, but you have to switch it on. Until you do, an order confirmation opened two days before a customer's next purchase can take the credit for it.Every other channel claims the same order. Klaviyo marks its own homework, as Meta and Google do (we've covered why Meta Ads and Shopify numbers don't match). Each platform counts the orders it can connect to its own touch, by its own rules, and none subtracts what another has claimed.Gross against net. In Klaviyo's Shopify data reference, a Placed Order's value is the order total including shipping, taxes and tips, with discounts applied. Shopify's net sales leave out shipping and tax and take off returns.Refunds never come off. Klaviyo's own data reference puts it plainly: "Shopify subtracts canceled and refunded orders from their revenue calculation, while Klaviyo does not." Klaviyo records a full refund as a separate Refunded Order event, partial refunds don't register at all, and nothing done to an order after it first syncs reaches Klaviyo. Returns, exchanges and refunds to store credit all come off Shopify's numbers and leave Klaviyo's untouched. That matters more than most brands realise: in their 2025 survey, the NRF and Happy Returns found US retailers expected 19.3% of online sales to come back that year, and online apparel runs nearer a quarter. Consumables and beauty sit far lower, but rarely low enough to ignore.Dates move. Klaviyo's campaign and flow reports group conversions under the date the message was sent, in your account's time zone, so an order placed on the 3rd from a campaign sent on the 29th lands in last month's campaign numbers. Shopify reports it on the 3rd.Currencies may not match. Klaviyo doesn't convert currencies in its reports (Benchmarks aside), and suggests a separate account per region if you trade in several. If you sell in more than one currency, check what Klaviyo is adding together before you compare it with anything.Take an illustrative month for a $12m subscription brand, in round numbers. Shopify shows $1m of net sales. Klaviyo's dashboard shows $430,000 of attributed revenue across email and SMS, 43% of the business, and that's the figure in the retention team's update.Like for like. Tax and shipping add about 10% to order values, and 15% of what was ordered is later refunded or cancelled. On a net sales basis, Klaviyo's figure is closer to $330,000.Without renewals and transactional messages. $70,000 of the attributed revenue is subscription renewals and orders credited to transactional emails. Take them out and email is at $260,000.Clicks only. Preview the account with opens and SMS deliveries switched off, leaving clicks, and the figure drops to around $170,000, or 17% of net sales.Same month, same customers: $430,000 on the dashboard and $170,000 on settings we'd defend. Email may still be one of the best-value channels this brand has, and for a retention business it often is. The dashboard just says it's worth two and a half times more.Defaults are a starting point. These are the settings we'd begin from for a $3m to $50m retention brand.UTMs on, everywhere. In Settings > Other > UTM tracking, switch on automatic UTM parameters. It only applies to new campaigns and flow messages, so go back through your live flows and switch it on there too. UTMs are how Shopify, GA4 and every other tool see email at all.Email: clicks only, three days. Untick email opens. Apple's automatic opens have made them close to meaningless as evidence that anyone read anything, and Klaviyo's own guidance warns that leaning on opens biases attribution. Three days covers the customer who clicks on Friday and buys after the weekend.SMS: clicks only, one day. Untick deliveries. A delivered text tells you a phone received it.Exclusions on. Exclude interactions with transactional messages and Apple Mail Privacy Protection opens (a safety net if opens ever come back), and keep bot clicks excluded, which Klaviyo does by default.Renewals out. Klaviyo has no attribution setting for this, so build a custom metric: Analytics, Metrics, Create custom metric, based on Shopify's Placed Order, filtered to exclude orders whose Source Name starts with "subscription_contract", which is how Klaviyo defines its own non-recurring order metric. Then choose it as the conversion metric in your campaign and flow reports. Klaviyo creates this metric for you on its Personalization and Advanced KDP plans, and the standard plan allows one custom metric, which this is a good use of. If you work with an AI assistant connected to Klaviyo's official MCP server, it can create the metric for you. Something like: "Create a Klaviyo custom metric called Non-Recurring Order, aggregated by value, using Shopify Placed Order and excluding events whose Source Name starts with subscription_contract."Use Klaviyo's Compare model tool to preview each change before you save it, since you only get three saves in 30 days. Expect email's number to fall. That's the point.Tighter settings get you closer, but they're still a judgement call. The only direct measure of what email adds is a holdout: keep a random group of customers out of your messages and compare what they buy with everyone else.Be realistic about it. Small lifts need big samples to detect, and Klaviyo's global holdout, which pauses all messaging for a random slice of your list, needs at least 400,000 profiles before you can switch it on. A flow's random-sample split gives you a control group for a single flow, but most brands in our range won't have the order volume to read anything meaningful from it in a sensible time. Nobody running a lean $20m business is going to repeat this every quarter.If your list is big enough, run one holdout across the board about once a year and treat it as calibration. Adjust the attribution windows until Klaviyo's reported revenue sits close to the lift the holdout measured. It will never match exactly, because your mix of email, SMS and campaign types and the season you tested in will drift, but it puts email in the right ballpark, where the defaults can report two, three or four times what it earns. Marketing mix modelling is the other route, and for most brands our size it's out of reach on cost and complexity.Klaviyo's attributed revenue, on tightened settings, is for comparing messages and flows against each other. As a ranking it's useful even when the total is high.Shopify's marketing reports default to last non-direct click. That doesn't measure what email adds any more than Klaviyo does, and it favours channels close to the purchase, like email, over ones further back, like Meta or YouTube. But its channels add up to the revenue you took, which is worth a lot when every other tool's total is bigger than the business.A calibrated Klaviyo number, if you can run the holdout, is the one to take into budget and headcount decisions for retention.Blended CAC and LTV:CAC are the numbers to run the business on. No platform's attribution settings can inflate them.Back to the year-end meeting. Every team's number was real, and the business still went backwards, because nobody was measuring the same thing. Marketing has the same problem every month. When the retention team's email figure, paid social's conversions and Google's brand search credit add up to more than Shopify took, finance trusts none of them, and decisions stall while each team defends its dashboard.Attribution can't be solved, so the most effective thing a small or mid-sized brand can do is agree a shared view of the world with sensible defaults: Shopify as the ledger, tighter platform settings, a calibration when you can afford one, and sources cited every time. That's what a company brain for D2C is for. We rebuild one common view of the numbers from structured data across every system you run, read the business LTV:CAC first, and our agents act on that view in the open.If you're below roughly $1m in revenue, the samples are usually too small to separate signal from noise, and tightening your Klaviyo settings is the best place to start.If you'd like a second pair of eyes on your Klaviyo setup and why your numbers disagree, book a call at signalovernoise.app/book-a-call or drop us an email at hello@signalovernoise.app.