Real-world application: email & CRM marketing that actually drives revenue

Black Friday 2015 knocked Gymshark's site offline for the best part of eight hours, at the highest-traffic moment of its trading year. The infrastructure answer was a move to Shopify Plus. The harder question arrived afterwards: a brand that had built demand through YouTube athletes and Instagram was renting its audience, and the only asset it owned outright was the email and SMS file sitting on top of its Shopify data. This lesson follows one DTC lifecycle programme of that shape end to end, on that stack: three flows, what each one is worth, and where each one breaks.

Core concept: one stack, three flows, one revenue line

The commerce layer is Shopify Plus: orders, refunds, inventory by variant, customer records, checkout events. The lifecycle layer is Klaviyo, the ecommerce ESP that sells exactly this integration (Shopify invested in Klaviyo in 2022; Klaviyo listed on the NYSE in September 2023). The connection runs both ways. Shopify pushes viewed product, added to cart, started checkout, placed order and fulfilled into the persistent profile the foundations lesson describes; Klaviyo writes back engagement, consent state and predicted next order date so segments stay current without anyone exporting a CSV.

Three flows carry most of the automated revenue in a programme like this: welcome, browse abandonment, win-back. In Klaviyo's published benchmark data, automated flows take a small single-digit share of total sends and around a third of email-attributed revenue. That ratio is the entire argument for building flows before campaigns.

Key sub-concept 1: the welcome flow and what its discount costs

Entry is the on-site pop-up: 10% off the first order in exchange for an email address, with SMS consent as a second step. Welcome flows carry the highest revenue per recipient of anything except checkout abandonment, and the first email routinely opens north of 40% because intent is fresh.

Now price it. On a £45 average order at roughly 65% gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition →, a 10% code hands back £4.50, most of it to shoppers already on their way to checkout. The margin line, not the revenue line, decides whether the pop-up is working. A no-discount variant (early access to a drop, size guidance, restock alerts) grows the file more slowly and converts less, but every order lands at full price. Test the two against contribution margin per new profile, not against sign-up rate. The other cost is behavioural: a discounted first order teaches a customer to wait for a code before the second one, and that habit shows up months later as a softer full-price repeat rate.

Key sub-concept 2: browse abandonment, gated by inventory

Browse abandonment only fires for a session the profile can be matched to, which is a minority of traffic for most DTC brands: a click from an earlier email, a logged-in account, an identified device. Everyone else is invisible, so the flow's ceiling is set by identification rate long before copy matters. Trigger delay of two to four hours works for apparel, with the suppression rules the frameworks lesson sets out keeping anyone who reached checkout or bought in the last few days out of it.

The failure mode specific to apparel is size. The product page still exists, the medium the shopper was looking at is gone, and the email drives a click straight into disappointment. If a third of your size variants are out of stock in peak weeks, roughly a third of your browse-abandon sends are pointing at something nobody can buy. Gate the send on the Shopify inventory level for the variant viewed, and route the rest into a back-in-stock flow instead, which converts better than the browse email ever did.

Key sub-concept 3: win-back, and the incrementalityincrementalityThe share of results (sales, conversions, revenue) that only happened because of a marketing action, not what would have occurred anyway.View full definition → problem

The 90-day lapse window is inherited folklore. Pull the median gap between first and second order out of Shopify order history and set the trigger against that. If the median is 150 days, a 90-day win-back is paying discounts to people who were coming back anyway, and the flow reports the revenue as its own.

Klaviyo's data puts win-back recovery on lapsed segmentssegmentsDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → in the 5 to 15% range. That headline is not the number to manage. Hold back 10% of each win-back cohort for a quarter, send them nothing, and compare purchase rates. Expect the incremental figure to be a fraction of the reported one. Sequence matters too: leading with the deepest discount caps everything that follows, so open with a new drop or a restock in their size and hold the code for the final message.

Key sub-concept 4: whose number counts, shopify's or klaviyo's

Klaviyo's default attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition → credits an order to a flow if the shopper opened or clicked within a few days of buying. Meta and Google claim the same order on their own windows. Add the channel reports together and the total comfortably exceeds what Shopify actually banked, often by a wide margin. Pick one number of record before you build the dashboard: Shopify's own order data for anything reported upward, ESP attribution for flow-level optimisation only. Teams that skip this argument spend the following year defending revenue that never existed.

Real-world case 1: what each flow contributed

Work it with round numbers for a list in the low millions. Welcome: 500,000 new profiles a year at roughly £1.50 revenue per recipient is about £750,000, the highest per-head return of the three. Browse abandonment: far more sends, maybe 2 million eligible triggers, but at perhaps 20 to 30 pence per recipient it lands in the same range on volume alone. Win-back: the smallest audience of the three, a few hundred thousand lapsed buyers a year, high revenue per recipient because the basket is a full order, and the most exposed to the holdout test above.

Two things follow. The welcome flow is the one to protect, because it is the only flow whose audience you control the size of. And cart abandonment, which sits outside this trio, will usually out-earn all three per recipient, so a programme that builds browse abandonment before checkout abandonment has sequenced itself badly.

Real-world case 2: black friday, when the flows should be turned down

During a sitewide 40% weekend, the win-back email offering 15% off is worse than useless: it burns margin on a customer who was already getting a better deal, and it tells the brand's best-informed buyers that the flow logic is asleep. Same for the 10% welcome code, which reads as an insult next to the public offer.

The peak-week routine on this stack: swap flow creative for the promotional message rather than pausing the flows outright, stretch the browse-abandon delay so it does not collide with campaign sends, and suppress every discount-bearing automation for the duration. Complaint rates spike during peak because campaign frequency spikes, and damage to a sending domain takes weeks to repair, which is precisely the weeks that carry the year's revenue.

Real-world case 3: the pop-up that grew the wrong list

Turn the pop-up on for every session on every page and the file grows fast. It also fills with giveaway hunters who never open anything. Since February 2024, Google and Yahoo have required bulk senders (above 5,000 messages a day to their users) to authenticate with SPF, DKIM and DMARC, provide one-click unsubscribe, and keep spam complaints below 0.3%, which is three complaints per thousand delivered. A padded list walks straight at that line, and the first thing to suffer is the transaction-adjacent mail that actually earns money.

The fix is unglamorous: a sunset rule that suppresses anyone who has not opened or clicked in 90 to 120 days, and acceptance that the file will shrink. Campaign revenue dips in the first month because you are mailing fewer people. It usually recovers within a quarter, because the people still receiving mail are the ones landing in the inbox.

CMO action items

• Put a permanent 10% holdout on the win-back flow and report incremental revenue, not attributed revenue, in the monthly pack. Do the same for any flow whose main lever is a discount
• Name the number of record this quarter. Shopify order data upward, ESP attribution downward, and no dashboard that mixes them
• Check that browse abandonment is gated on variant-level inventory, and that the welcome and win-back discounts are suppressed for the whole of your peak promotional window

Common mistakes that kill results

• Judging the welcome flow on revenue per recipient while ignoring the margin the discount gives away. The flow can look like your best performer and still be your least profitable acquisition path
• Building browse abandonment before checkout abandonment because it feels more sophisticated. The shopper who reached checkout is worth several times the shopper who looked
• Treating flow revenue as found money. Some of it is customers who would have bought anyway, and the only way to know how much is to stop mailing a slice of them and watch

Key takeaways

• Flows earn a disproportionate share of email revenue off a tiny share of sends, which is why they come before the campaign calendar
• The welcome flow's real metric is contribution margin per new profile, not revenue per recipient
• Browse abandonment lives or dies on identification rate and variant-level stock; ungated, it sends people to a size that no longer exists
• Set the win-back trigger off your own median inter-purchase gap in Shopify, then measure it against a holdout, because the headline recovery rate flatters the flow
• Peak weeks need flows turned down, not up: a 15% win-back code during a 40% sitewide sale burns margin and credibility at once
• Growing the file faster than engagement puts you near the 0.3% complaint ceiling Gmail and Yahoo have enforced since February 2024, and the mail you lose first is the mail that pays