Same Revenue Target. Fewer Clicks. Something Has to Give.

US retail is having a good year. The National Retail Federation expects sales to grow 4.4% in 2026, to $5.6 trillion, against a 3.6% average over the last ten years excluding the pandemic period from 2020 to 2022. The money is out there, and there is more of it than usual.

Your session count probably doesn’t reflect that. For most enterprise retailers the top of the funnel has gone flat or slightly negative while the revenue number on the plan went up. That combination is new, and it breaks the move every ecommerce team has reached for since 2010.

For fifteen years, traffic behaved like a faucet. Revenue gap in Q3? Open it wider. Buy more clicks, bid harder, publish more pages. The lever was expensive but reliable, and it meant conversion work could stay a nice-to-have: a thing the team got to after the campaigns shipped.

That faucet is being closed from both ends at once. The interesting part isn’t the closing. It’s what it does to the math on the lever you haven’t been pulling.

The Old Math Had Two Variables. One of Them Just Got Expensive.

Old math: more visitors × same conversion rate = more revenue.

New math: same visitors × better conversion rate = more revenue.

Same equation. The only thing that changed is which variable you can still afford to move, and that single swap rewrites what your team should be doing on a Tuesday.

Search Still Sends Traffic. It Sends Less of It Per Impression.

Pew Research Center tracked the actual browsing behavior of 900 US adults across nearly 69,000 Google searches. When an AI summary appeared in the results, users clicked through to a traditional search result on 8% of visits. When no summary appeared, they clicked on 15%, close to twice as often. They clicked a link inside the summary itself on 1% of visits. And they ended the browsing session entirely on 26% of pages carrying a summary, versus 16% without one.

One caveat worth stating plainly: that study covers general Google search, not retail queries specifically, and it captured browsing in March 2025. Commercial queries behave differently from informational ones. Treat it as direction and magnitude, not as your category’s exact number.

The direction is not in dispute. Search can now send you fewer clicks from the same visibility, and nobody on your team has to do anything wrong for that to happen.

There is a second half to that story. AI-referred visitors convert better than any other channel, while most retail product pages remain unreadable to the systems sending them. We covered that separately, and it is a real opportunity. This post is about the arithmetic underneath your plan.

Buying the Clicks Back Is a Margin Decision Dressed as a Growth Decision

The obvious response is to pay for what you used to earn. Everyone had that idea at once, which is exactly the problem.

Tinuiti’s Q2 2026 benchmark report puts Google paid search spend up 14% year over year and Google Shopping up 18%. Amazon Sponsored Products rose 38%, or 23% once you strip out the Prime Day calendar shift. More money is chasing those clicks. They haven’t become any easier to win.

So the spend is real, and some of it is necessary. Be honest about what it buys. Replacing organic clicks with paid ones holds your session count steady while your cost per acquired session climbs. Volume flat, margin down. That shows up in a board deck as "we maintained traffic," which is true, and as a quietly worse contribution margin, which is also true and much less likely to get said out loud.

There is a version of this that is the right call: a category where you genuinely cannot afford to cede share this year. Fine. Just book it as defense, not growth, and don’t let it consume the budget for the lever that still has room in it.

Two Out of Three Ecommerce Checkouts Test as Mediocre or Worse

Baymard Institute’s aggregate of 50 separate studies puts average documented cart abandonment at 70.22%, as of its September 2025 update. That number is famous enough to have gone numb.

This one hasn’t. Across 344 top-grossing US and EU ecommerce sites, Baymard found that 65% score mediocre or worse on checkout usability, and puts the potential conversion improvement available from better checkout UX at roughly 35% for the average large-scale ecommerce site.

Sit with the asymmetry. Your team is fighting for single-digit percentage improvements in an auction where the price goes up every quarter, while roughly a third of the achievable conversion on traffic you have already bought sits unclaimed at the bottom of the funnel. Not theoretically available. Estimated as achievable, across hundreds of sites, by the group that has studied this for fifteen years.

Two out of three enterprise sites are running a checkout that tests as mediocre or worse. Yours might be the third. It’s worth knowing which, and most teams don’t, because nobody has looked recently.

Point It at the Money That’s Furthest Down the Funnel

When traffic is flat, conversion effort stops being a general virtue and becomes an allocation problem. Most teams allocate by whoever asked loudest, or by whichever page the last consultant flagged. There’s a better rule, and it’s a single line.

Rank by acquisition cost already sunk × measured friction.

A session that reached your cart cost you everything the click cost, plus everything it took to get that shopper through discovery, a category page, and a product page. That session is the most expensive inventory you own, and it is one interaction away from revenue. A session that bounced off a PLP cost you the click and nothing more.

Both are worth fixing. They are not worth fixing in the same week.

This is uncomfortable in practice, because the bottom of the funnel is where the politics live. Checkout touches payments, fraud, tax, and three teams who each have a legitimate reason to say not this quarter. The top of the funnel is easier to change and easier to show in a deck. So effort drifts upward, away from the money, which is precisely the drift a flat-traffic year can no longer afford.

Run a shared checkout template across brands and the drift becomes structural. The fix that would move the most revenue is also the fix that needs the most signatures.

Why Conversion Programs Stall, and It Isn’t Ambition

Every ecommerce leader reading this already knows conversion is the answer. The list of things they’d fix isn’t short. So why does the list survive contact with the calendar?

Two reasons, and they compound.

The first is resolution. You know cart abandonment is 70%. You don’t know that it spikes on mobile at the shipping-options step for returning customers in one region, because that cut isn’t in the standard reporting package and getting it means a tagging plan, a ticket, and an analyst with their own queue.

The second is release cadence. Even once you know, the fix waits for a window. We’ve broken that wait down elsewhere: three of the four delays between an insight and a live change happen before a developer ever sees the ticket.

Those are the Insight Gap and the Activation Gap, and they multiply rather than add. Slow sight plus slow hands means the one lever with room in it moves a few times a year instead of a few times a month. A program moving at that pace never accumulates enough clean results to prove what it earned.

No amount of alignment gets a merchandiser a checkout variant without a release window. That is architecture, not coordination, and it is the problem Fastr Workspace exists to collapse: one place where finding the friction and shipping the fix are the same workflow rather than two systems with a handoff between them. Behavioral intelligence that surfaces checkout friction without a tagging plan, and publishing that puts the fix live without a development ticket.

Four Places This Argument Doesn't Hold

Any VP who has run a conversion program will have already thought of them.

Your conversion gain might not be incremental either. That objection came from Rebecca Kerper, former Chief Digital Officer at HSN (on the RTM Nexus panel): defaulting to conversion rate settles nothing, because the number might be incremental or it might be the number you were going to get anyway. It is the same skepticism this post applies to paid clicks, and it has to point both ways. The practical answer is the one that applies to any test: hold traffic back, and measure the winner against a control rather than against last month.

Conversion isn’t infinitely elastic. Baymard’s 35% is a ceiling for sites with real checkout problems, not a forecast for yours. A site already scoring well has far less to claim, and the second 5% costs considerably more than the first. Anyone promising you a compounding conversion curve forever is selling something.

Some traffic declines aren’t a search problem. If sessions are down because the assortment got stale, a competitor undercut you, or the brand lost relevance with a cohort, conversion optimization is a painkiller applied to the wrong limb. Check that the traffic you’re losing is traffic that would have bought.

Flat traffic isn’t permanent by decree. Retail media, marketplaces, owned channels, and AI-referred visitors are all real routes back to volume. The argument here isn’t that acquisition is dead. It’s that acquisition is no longer the cheap lever, and a plan that still treats it as the default is allocating last decade’s budget to this decade’s market.

What Changed Isn’t the Target

The plan didn’t get more aggressive. The mechanism you’ve used for fifteen years to hit it just repriced.

Every point of conversion you find is now worth what an entire acquisition campaign used to be worth, and unlike the campaign, you only have to win it once. It keeps paying on traffic you have already bought, on sessions you are already getting, in a market NRF expects to grow 4.4% in 2026.

The teams that come out of this year ahead won’t be the ones who bid better. They’ll be the ones who turned the traffic they already have into more revenue than anyone else could.

Want to know which third your checkout is in? We’ll show you your own site, not ours. 20 minutes. No slides.

Book a demo

Frequently Asked Questions

How do you grow ecommerce revenue when traffic is declining?

By shifting budget from acquiring new sessions to converting the ones you already pay for. Organic click-through has compressed while paid costs have risen, so conversion is now the cheaper lever. Rank fixes by acquisition cost already sunk multiplied by measured friction: the sessions furthest down the funnel are the most expensive inventory you own.

How much conversion lift is realistically available from checkout optimization?

Baymard Institute benchmarked 344 top-grossing US and EU ecommerce sites and found 65% score mediocre or worse on checkout usability. It puts the potential conversion improvement from better checkout UX at roughly 35% for the average large-scale ecommerce site. That is a ceiling for sites with real problems, not a forecast for every site.

Are AI summaries reducing clicks to retail websites?

Pew Research Center tracked 900 US adults across nearly 69,000 Google searches and found users clicked a traditional result on 8% of visits when an AI summary appeared, versus 15% when one did not. The study covers general search rather than retail queries specifically, and captured March 2025 browsing, so treat it as direction and magnitude rather than a category-specific figure.

Is it worth buying paid search to replace lost organic traffic?

Sometimes, but book it as defense rather than growth. Tinuiti’s Q2 2026 benchmark shows Google paid search spend up 14% year over year and Google Shopping up 18%. Replacing organic clicks with paid ones means paying for traffic you used to earn: volume defended, margin spent.

Why don’t ecommerce conversion programs ship the fixes they identify?

Because finding the friction and fixing it happen in different systems. Friction is visible only at a resolution that requires tagging plans and analyst queues, and the resulting fix waits for a release window. Both delays are architectural rather than a failure of prioritization or alignment.