Case study: when store traffic falls 3x, what email can and cannot do

The question the client asked

“Why were the results so much better before?” Fair question. In January to April 2025 Klaviyo attributed roughly twice as much revenue to email per month as it does in 2026. The store itself is bigger: October 2026 sales were about 30% above the comparable month before. Same agency, same team, same Klaviyo account.

We pulled 22 months of campaign, flow, form and metric data through the Klaviyo API to answer it. This is what the data said.

Finding 1: the popup never got worse. The traffic did.

The signup popup converted between 4.4% and 5.5% of unique viewers every single month of the period. What changed is how many people saw it: about three times as many in early 2025 as in 2026. Three times fewer people entered the funnel, so three times fewer new subscribers reached the welcome flow, and the welcome flow is where a new subscriber’s first order comes from.

Nothing in Klaviyo can fix that. The traffic came from paid social and from a creator audience that moved on. Email converts the audience you bring it; it does not create one.

Finding 2: early 2025 was propped up by imported lists.

In the strong months the account was adding two to four times more subscriptions than the popup produced. The difference was imported lists from older stores and partner lists. An imported list gets a one-time welcome send and buys once; it does not come back. The reachable audience shrank by about a third as those imports churned out. The early numbers were real revenue, but not a baseline anyone could keep.

Finding 3: discount-only campaigns wore the list down.

Campaign click rate was 0.5% to 0.6% in early 2025 and 0.35% to 0.4% in 2026, on an account that sends more than a million campaign emails a month. Almost every send carried a sitewide discount. A list that gets a discount every week stops clicking on anything else, and the people who do click are the ones who would have bought anyway. The campaign share of email revenue fell while flows (65% of email revenue over the year) held up, because flows respond to behaviour, not to the offer of the week.

What email still did

Over the 12 months to October 2026 email attributed 22.4% of store revenue: exactly the benchmark median for Shopify stores in our portfolio, and 17% to 19% in the most recent months. The flows (welcome, abandoned checkout, browse abandonment, post-purchase, winback) do two thirds of it. Flow click rate is 2.88% and conversion 0.96%, both above the benchmark. The program is healthy; the funnel above it is smaller.

What we changed

1. Reporting now shows the client popup views, submits and new subscribers next to email revenue, so a traffic drop is visible the month it happens, not a year later.
2. Campaigns moved from the whole list to a 90-day engaged segment with recent buyers suppressed, and from discount-only to a mix of product stories and one offer a month. Click recovered on the first product-led sends.
3. The non-product messages that had been in the campaign calendar moved into their own flow, so campaign cadence dropped from 3 to 4 a week to 2.

What to take from it

• When email revenue falls, look at the top of the funnel first: popup views, submits, new subscribers. If conversion held and views fell, the answer is traffic.
• Imported lists make a great quarter and a bad baseline. Report them separately.
• A weekly discount is a loan against the list. The click rate pays it back.

The store is on our Pro retainer. The free 30-minute audit includes exactly this funnel read: views, submits, subscribers, flow and campaign share.