When Discovery Happens Before the Click: How AI is Reshaping Ecommerce Performance Marketing

Performance marketing has traditionally been built around the idea that discovery starts when someone clicks an ad or types a query into a search engine.

That’s no longer a given.

Increasingly, more customers are becoming familiar with brands through AI-led discovery tools and conversational platforms – forming opinions, evaluating options and building shortlists before they ever click a paid ad or open Google.

Note that this isn’t about AI replacing paid media – it’s about the customer journey shifting, and performance strategies needing to evolve to match how decisions are now made.

This article explores exactly how AI-led discovery is changing the customer journey and the role of paid media within it. We’ll look at the growing importance of validation in paid ads, and how teams can adapt to optimise for this. There’s also a nod to Google and Shopify’s Universal Commerce Protocol (UCP) announcement, including the Direct Offers ads pilot, and how these may start to impact the customer journey further.

AI discovery starts with questions, not keywords

The early stage of the customer journey used to be powered by keywords: a shopper has a need, translates it into a search query, and evaluates the results that follow. Whilst that model still matters, it’s no longer the only starting point.

AI-led discovery tools have introduced a different behaviour: customers begin with context, using conversational prompts to describe their situation and get guided, summarised answers.

Instead of searching with keyword fragments, customers are increasingly kicking off discovery with complete questions. They’re asking ChatGPT and similar platforms questions like:

• “What are the best running shoes for overpronation?”
• “Which mattress brands are similar to Simba but cheaper?”
• “Is brand X legit? Where do they ship from and what’s the return policy like?”

This shift is reshaping discovery in three meaningful ways:

• It becomes conversational. Natural language questions are replacing keyword shorthand, making intent richer and more nuanced.

• It becomes comparative by default. Multiple brands are being evaluated in a single interaction, with pros, cons and trade-offs surfaced early. Customers expect recommended options tailored to their personal circumstances.

• It becomes validation-led. Reassurance, credibility and proof are becoming part of the discovery process itself, not something that comes later.

As a result, AI is increasingly influencing which brands are surfaced, which feel credible, and which make consumers’ shortlists.

The customer journey has shifted upstream

Performance marketing has traditionally been optimised around a fairly linear path:

Keyword → Click → Consider → Buy

This is a model that lends itself well to measurement, and it’s shaped how many teams structure budgets, campaigns and KPIs.

AI-led discovery changes this journey, moving the “consideration work” earlier in the flow:

Question → Compare → Shortlist → Validate → Click

Now, by the time users reach paid search or paid social, they are often:

• aware of their options in the category landscape, with a pre-formed shortlist;
• carrying expectations about price, quality or delivery;
• actively evaluating trust and legitimacy; and
• looking for signals that confirm they’re making the right choice.

This is the key shift: the click is no longer the start of discovery. Paid media is increasingly meeting users later in the decision cycle, when the job is less about introducing a brand and more about building confidence in choosing it.

Why paid performance can feel harder than it used to

When customers have already compared and shortlisted, paid media operates in a tighter race – and the brands already “trusted” by the user start with an advantage.

This is easy to overlook when performance is measured purely on last-click attribution. If discovery and preference formation happens upstream – via AI-led tools, communities, creators, reviews and recommendation engines – paid search can end up taking credit for decisions that were already leaning one way, while simultaneously being judged for not creating demand from scratch.

For many ecommerce marketing teams, that mismatch shows up as increasing pressure on efficiency: more expensive auctions, less headroom, and a smaller pool of users who are truly ‘discovering” through ads.

Paid search has become a validation layer

Paid search remains a critical channel in ecommerce, but its role is shifting for many brands.

In an AI-influenced journey, paid media performs best when it does three things well:

• reinforces trust and credibility;
• aligns with what customers already believe; and
• removes friction at the point of decision.

Ads that validate existing intent tend to outperform ads that try to create it from scratch. This means that performance marketers need to rethink what they optimise for:

1. From attention to assurance

If a user already has a shortlist, your ad has a narrow window to answer the question “Why you and not them?” – and quickly.

Specificity is key here. Generic claims struggle in validation environments because they don’t reduce perceived risk. Strong ads tends to be anchored in tangible signals, such as:

• clear, concrete value propositions (not vague superlatives);
• delivery and returns clarity;
• meaningful social proof points (like reviews and ratings);
• authority signals, such as certification, awards and partnerships; and
• pricing transparency (or a confident reason for premium positioning).

2. From clever copy to credible proof

In a validation phase, the user is scanning for risk reduction. As such, your creative, extensions and landing pages should make reassurance frictionless and immediate:

• “Free UK delivery over £X”
• “30-day returns”
• “Rated 4.8/5 by 14,000 customers”
• “Next-day dispatch”
• “Official partner / accredited / clinically tested”

These details might feel operational, but they can often be the deciding factor when someone is choosing between two credible options.

3. From channel-level optimisation to a joined-up narrative

Validation doesn’t happen in a single place. AI results, paid ads, product pages and on-site content all need to tell the same story. This helps to reduce cognitive friction – and conversion becomes easier to scale.

If each channel shares a different narrative, the click becomes more expensive and less productive, even when targeting and creative are strong.

The impact of Direct Offers in Google Ads

We’re particularly interested to see how the launch of Google’s Direct Offers pilot impacts the role of paid media in the online shopping journey. With this new AI-powered initiative, select retailers will be able to present exclusive deals directly in AI Mode – which means that when shoppers are ready to buy, brands can meet them with the right offer, at the right time, right inside the conversation.

Google has outlined how this will work in practice:

“Imagine you search ‘I’m looking for a modern, stylish rug for a high-traffic dining room. I host a lot of dinner parties, so I want something that is easy to clean.’ Google already elevates the most relevant products to meet your search criteria. But often, you are only ready to buy if you’re getting a great deal. Now relevant retailers have an opportunity to also feature a special discount. This helps you get better value and helps the retailer close the sale.”

The pilot is expected to expand to support the creation of other ad-based offers beyond discounts, such as bundles and free shipping – further helping to validate intent and provide immediate assurance.

Performance marketing is a connected system

In modern ecommerce, performance doesn’t sit neatly inside one channel. It’s an ecosystem of connected systems working together, where discovery, demand capture and conversion should all reinforce one another.

A useful way to think about it is as a flow of confidence:

• AI discovery shapes perception and awareness, framing expectations
• Paid media validates demand and supports choice
• Your Shopify experience converts trust into revenue

Rather than replacing performance marketing, AI accelerates the performance flywheel. To keep that flywheel spinning smoothly, every stage needs to support the next. Leakage anywhere in the system can reduce efficiency everywhere.

For instance, if discovery creates interest but your ads feel generic, you lose intent before reaching your site. If ads drive clicks but your product pages don’t answer validation questions, you lose conversion at the point of decision. And if UX converts but the post-purchase experience disappoints, you lose lifetime value – and your CAC pressure returns as you’re forced to keep reacquiring customers.

This is why the best performance outcomes are increasingly tied to cross-functional alignment, not just media optimisation. When messaging, merchandising, UX and retention are pulling in the same direction, the system compounds: paid media becomes more efficient, conversion improves and customer value grows.

It’s worth noting here that, depending on adoption, new agentic commerce initiatives like Google and Shopify’s UCP could start to change what conversion looks like and where it takes place. Soon, consumers will be able to purchase from eligible retailers as they’re researching in Google, with a new checkout feature on product listings in AI Mode in Search and the Gemini app. This could result in the customer journey moving even further upstream, minimising the need for high-intent shoppers to visit retailers’ ecommerce sites.

What leading brands are doing differently

The brands succeeding in AI-led discovery environments are strengthening the fundamentals that both AI systems and consumers reward: clarity, credibility and consistency. In practice, that means making it easy for machines to interpret the brand – and easy for people to trust it – at every stage of the journey.

Here’s what that looks like.

Creating content that AI can understand & confidently reference

AI-led discovery tends to surface information that is easy to parse, compare and summarise. Content therefore performs best when it’s:

• structured and specific (with clear headings, clear claims and clear answers);
• clear, concise and authoritative (no vague marketing language);
• consistent across sources (site, reviews, FAQs, feeds etc. all reinforcing the same facts); and
• written in a way that maps to real customer questions.

When done well, helpful, evergreen content becomes a growth asset, acting as a discovery and validation engine. Typical examples include:

• buying guides built around use cases and concerns;
• comparison pages;
• FAQs that answer questions clearly; and
• product pages that explain why a product is right, not just what it is.

Strengthening entity signals across your brand presence

Your brand isn’t just your website – it’s the sum of signals across the web and your owned channels. These signals increasingly shape how your brand is represented in AI-led discovery.

Leading brands focus on consistency across:

• product naming, taxonomy and categorisation;
• brand story, positioning and differentiators;
• policies (e.g. delivery, returns, warranty);
• reviews and reputation markers; and
• structured data and merchant feeds.

When these elements align, it becomes easier for both people and machines to understand what you offer. When they don’t, AI systems (and customers) are left to fill in the gaps, which is rarely helpful for conversion.

Using paid ads to reinforce credibility, not chase first impressions

Awareness still matters. But when discovery starts with context, paid media works best when it mirrors the reality of how shoppers are thinking and deciding.

In practice, we’re seeing stronger performance when teams shift towards:

• more emphasis on proof points and specifics in ad copy;
• creative that reflects actual customer questions, objections and decision criteria;
• landing pages designed for decision-ready users (rather than first-time browsers);
• better segmentation between “educate” and “convert” intents.

Optimising Shopify UX for informed, decision-ready users

AI-led discovery can send you higher-intent traffic – but also more demanding traffic. These users arrive with sharper expectations and less patience for ambiguity. They want clarity fast.

Top-of-mind questions for them may include:

• What makes this product right for me?
• Why is it priced this way?
• Can I trust the quality?
• What happens if it doesn’t fit / suit me / work?
• When will it arrive?

This is where Shopify UX becomes a performance lever. Leading brands design product and category experiences to resolve validation questions quickly through:

• strong product page structure, including benefits, specs, proof, FAQs, policies etc.;
• clear navigation that serves “I know what I want” shoppers;
• comparison tools, bundles and guided selling where relevant; and
• speed and mobile performance (because validation is impatient).

What this means for ecommerce growth teams

Performance marketing is no longer only about capturing intent. It’s increasingly about supporting decisions already in motion – and making sure every touch point reinforces the customer’s confidence.

Teams that think in connected journeys rather than isolated channels are better positioned to build long-term brand credibility, improve paid efficiency over time, and create more resilient ecommerce growth strategies.

It also changes how success should be measured. If discovery and preference formation is happening before the click, you’ll likely need a broader view than last-click alone. That doesn’t mean abandoning performance rigour – the answer is to widen the lens and strengthen the signals you use to make decisions, including:

• understanding what paid is genuinely adding;
• better segmentation of intent (separating new vs returning, informed vs browsing and validation vs exploration behaviours);
• on-site engagement signals that show “pre-conversion confidence”; and
• post-purchase feedback loops that capture why a customer chose you – and what almost stopped them.

Ultimately, the goal of profitable growth remains the same – the path to achieving it is just changing. As discovery moves upstream, it’s about optimising the whole decision journey rather than just the last click.

How Swanky supports modern performance marketing

At Swanky, we help ecommerce brands adapt to changing customer behaviour by connecting AI discovery, paid media and Shopify performance into one cohesive strategy. Rather than treating channels in isolation, our digital marketing team focuses on building clarity, trust and alignment across the full customer journey – from first discovery through to conversion and repeat purchase.

If you’re curious how this shift applies to your performance strategy or upcoming priorities, please get in touch to speak to our friendly team of digital marketers.