Ads to a chat vs a landing page: what changes after the click

A landing page shows everything at once and waits. That works when the shopper already knows what they want. It struggles when they have a question the page doesn't answer up front: which size, which model, whether it fits, what shipping costs, whether returns are easy.

The research on that gap points the same way:

• 70.22% of online shopping carts are abandoned, on average across 50 studies (Baymard Institute, updated Sept. 22, 2025).

• Cost, delivery and trust come up most. Among U.S. shoppers who abandoned for a reason other than "just browsing", 40% named extra costs, 20% slow delivery and 19% not trusting the site with their card (Baymard Institute, U.S. shopper survey, same page).

• Product pages are weakest on phones. Baymard rates 62% of mobile ecommerce sites' product pages "mediocre or worse", and in its testing, shoppers left products that suited them because of problems the page could have fixed (Baymard Institute, Product Page UX 2026, Mar. 18, 2026).

None of those figures describes Clickless. They describe the page an ad usually lands on.

The chat isn't magic. It moves the answer earlier. A shopper who would have left over a missing answer can ask for it.

One brand that sells a $400 product, and is affiliated with Clickless, ran a two-week test inside a single Meta campaign: it pointed 4 ad sets at its Product Expert and left the campaign's other ad sets on its site.¹

From October 10 to 23, the same Product Expert ad sets had 3.8 times the add-to-carts per ad dollar: 109 add-to-carts at about $20 each, against 15 at about $77 for the campaign's other ads.² Read that number with care. The Product Expert's handoff places the product in the cart, so an add-to-cart there isn't the same act as one on a product page. What it gives you is a bigger pool of cart-adders for retargeting and cart emails, which reach only the shoppers a store can contact.

This was not a randomized test. The two groups of ad sets can differ in creative, audience and timing, and the Product Expert ad sets ran for about two weeks, in October, while the campaign's report covers Aug 23 to Oct 24. Results vary. The full numbers and caveats are on the results page.

As of October 2026, we haven't found a public, randomized test that compares a chat with a landing page for the same ad. Most published "chat lifts conversion" figures come from vendors and measure shoppers who chose to open a chat, who were likely closer to buying anyway. That's why a 50/50 test on your own ads matters more than any vendor figure, ours included.

One related data point supports handing the shopper to the brand's own checkout rather than finishing the sale inside a chat. Walmart said purchases completed inside ChatGPT converted at one-third the rate of purchases where shoppers clicked through to Walmart's site (Walmart EVP Daniel Danker to WIRED, as reported by Axios, Mar. 24, 2026). On March 24, 2026, OpenAI said it would let merchants use their own checkout experiences while it focuses on product discovery. A Product Expert answers in the chat and lets your checkout do the checkout.

• One product, one reason to buy. An impulse product with nothing to explain often converts fine on a page.

• A page that keeps winning. If a page has beaten every challenger, test against it before you replace it.

• A form is the goal. If the ad exists to collect an email or a lead form, a page built for that may suit better.

• A long story to tell. Long-form pages with video and reviews in one scroll still have their place.

Many brands keep both: the page for some ad sets, the Product Expert for the products that raise questions.

1. Pick one ad set that already sells.

2. Duplicate it and change only the website URL, to your Product Expert's ad link. Keep the creative, audience, placements and budget the same.

3. Use Meta's A/B test, so each person sees only one version.

4. Run it for 7 days, one full click window, without touching the budget.

5. Compare cost per purchase on both sides, by Meta's count and by the orders confirmed in your store.

The full walkthrough is in How to split test the destination of a Meta ad.

You pay only when it sells, with no setup or monthly fee. There's no Clickless fee on sales in its first 7 days live, so a one-week test costs you the ad spend you'd run anyway. Clickless counts ad sales the way you already count Meta: 7-day click, never view-through.

Media buyers who bring a brand earn a share of Clickless's fee on it, paid out of Clickless's share. The brand's bill doesn't change.

Does sending ads to a chat lower cost per purchase?

It did for one brand: in a two-week test inside the same Meta campaign, about half the cost per purchase, $100 vs $203 on 24 vs 118 purchases (Meta-reported, Meta's default attribution; not a randomized test). Results vary, so run a 50/50 test on one ad set before you move budget.

Which ads can open a Product Expert?

Any ad that links to a web page: Meta website ads, and TikTok and Google ads. The same link works in emails, bio links and QR codes. Ads that open Instagram DMs are rolling out as Meta approves the app.

Do I need to build a new landing page?

No. You change the ad's website URL to your Product Expert's ad link. Each ad can have its own first line, deal and opening, set in the Product Expert.

Where do shoppers pay?

On your own checkout. The Product Expert hands them over with the product and the deal ready. Clickless never takes payment.

How are the sales counted?

Within 7 days of the ad click by default, Meta's standard click window, or 28 or 60 days if you choose. A sale after the window is shown but not billed, and a view without a click never counts.

• Media buyers: get your partner link and plan a 50/50 test at getclickless.ai/media-buyers.

• Brands: set up your Product Expert at getclickless.ai/brands. Free 7-day test.

Related: Clickless results · Split test a Meta ad's destination · Clickless vs on-site chat widgets · Run an ad to your agent (docs) · Pricing

One brand affiliated with Clickless, with a $400 product: a two-week test inside the same Meta campaign (report window Aug 23–Oct 24, 2026). Meta-reported, Meta's default attribution; 24 vs 118 purchases. Not a randomized test. Results vary.

One brand affiliated with Clickless, same Meta campaign, Oct 10–23, 2026. Meta-reported, Meta's default attribution; 109 vs 15 add-to-carts. The Product Expert's checkout handoff places the item in the cart. Results vary.

Meta is a trademark of Meta Platforms, Inc. Clickless isn't affiliated with Meta.