Ecommerce strategy: What brands need to rethink as the customer journey changes

The ecommerce customer journey is getting harder to map.

A customer might discover a product on social media, ask an AI assistant to compare their options, check reviews on a marketplace, visit a store, search the brand directly and finally buy through its website.

For brands, that creates a challenge. Ecommerce strategy can no longer focus solely on getting more people to a website and improving what happens once they arrive.

Discovery, decision-making and buying are becoming spread across more platforms, while AI is starting to sit between brands and customers at several points in the process.

But that doesn’t mean ripping up everything that worked before.

That was one of the clearest messages from our recent webinar, What’s Next for Ecommerce?, where Paul Casey of Glass Atlas was joined by Adam Pearce of Blend Commerce, Stephen Lloyd of RT7 Digital and Jason Gonsalves of Portas.

The technology is changing quickly. The principles behind a good ecommerce strategy are changing much less.

Alongside the webinar discussion, we’ve brought in perspectives from seven specialist agencies within the Pimento network, spanning SEO, paid social, CRO, ecommerce development, retail, customer experience, AI, data and retention. Together, they reflect the breadth of expertise brands can access through Pimento when an ecommerce challenge requires more than one specialism.

So, what should brands be thinking about?

In this article

• How customers find you

• Getting found in AI search

• Building better product pages

• Protecting your brand

• Connecting stores and ecommerce

• Getting more from your traffic

• Fixing website performance

• Making paid social work harder

• Using AI where it matters

• Building your ecommerce strategy

Discovery is no longer happening in one place

For years, ecommerce discovery was heavily centred around search engines, paid media and marketplaces.

Those channels still matter. But they are now part of a much wider picture.

Social platforms, creators, marketplaces, reviews, physical retail and AI assistants can all influence which brands and products make it onto a customer’s shortlist.

YouGov’s 2026 research into online discovery found that British consumers are already using AI assistants alongside traditional search, although trust remains a significant barrier to wider adoption.

DHL’s research points to a similar fragmentation. Its 2026 global ecommerce study found that 38% of shoppers are already using AI-powered chat or virtual assistants as part of buying, while social commerce continues to influence discovery and purchasing behaviour.

That doesn’t mean brands suddenly need to appear everywhere.

Stephen Lloyd of RT7 Digital raised exactly this issue during the webinar. As more marketplaces emerge, brands face a choice between expanding their presence and creating an increasingly complicated ecosystem to maintain.

Product information that is accurate in one place but outdated somewhere else is already a customer problem. As AI systems increasingly gather information from multiple sources, that inconsistency could become a discoverability problem too.

The objective shouldn’t therefore be maximum presence. It should be the right presence, with consistent and useful information wherever your customers are looking.

Ecommerce SEO is becoming a question of trust

That changing discovery journey also has implications for SEO.

Anthony Barone, Co-Founder of Studiohawk, argues that brands need to think beyond rankings.

Area of expertise: Ecommerce SEO & AI Search Visibility

Ecommerce SEO: Start Optimising for Trust

Here’s the thing about AI search, it doesn’t just crawl your site anymore, it judges it. Shoppers are asking LLMs to pick products for them, which means AI is doing the recommending on your behalf. If it can’t quickly work out what your product is, who it’s for and why it’s good, it’s not going to vouch for you. It’ll just move on to a brand that makes its job easier.

The good news is this isn’t really a new skill to learn, it’s mostly about tightening up the basics. Clean structured data, genuinely useful product descriptions and reviews that actually mean something.

How to actually get cited:

• Make your data machine-readable: schema markup is how AI models confirm what your product actually is, so treat it as non-negotiable, not a nice-to-have.

• Track citations, not just rankings: start monitoring how often you’re mentioned in AI answers, because that’s becoming the new visibility metric.

• Write for the follow-up question: if an AI can’t answer “will this work for X” from your page alone, it won’t recommend it with confidence.

Visibility used to mean ranking. Now it means being trusted enough to get quoted.

Your product pages have a bigger job to do

This creates an interesting tension.

For years, much of conversion optimisation has concentrated on removing friction. Reduce the steps. Simplify the page. Get customers to the call to action.

Now product pages also need to provide enough information for increasingly well-informed customers and the systems helping them make decisions.

Adam Pearce of Blend Commerce made this point during the webinar. Information about ingredients, materials, production, shipping and FAQs may once have been spread around a website. Increasingly, it needs to be readily available at product level.

That doesn’t mean cramming every possible piece of information onto every page. It means understanding the questions customers need answering before they feel confident buying.

The same principle applies outside traditional consumer ecommerce.

Ecommerce isn’t just a checkout

Stuart Bloodworth, Creative Director at Nice&Co, sees an opportunity for manufacturers that have traditionally dismissed ecommerce because their products aren’t suited to a simple online transaction.

Area of expertise: B2B Ecommerce & Manufacturing Websites

Most UK manufacturers don’t think ecommerce is for them.

Their products are too complex. Pricing varies. Orders need a conversation. So ecommerce gets filed under “online shop” and the website stays exactly where it is: a digital brochure.

That’s the mistake.

Ecommerce isn’t really about the basket. It’s about removing friction.

A good manufacturing website should help a buyer find the right product, understand where it fits, compare options, download the technical detail and build a shortlist. That shortlist can become an enquiry, complete with product references, quantities and requirements. No vague “please call me” form. No salesperson spending the first half of a conversation working out what the customer actually wants.

The same work improves reach. Clear product structure and useful, specific content give search engines and, increasingly, AI search far more to understand and surface. That means more opportunities to be found by buyers who may never have heard of the business.

For some manufacturers, the next step will be repeat ordering or customer portals. For others, the sale will always need a human being involved. That’s fine. Ecommerce doesn’t have to replace the sales process. It should make the sales process better.

UK manufacturers don’t all need an online shop. But they do need websites that work harder: bringing in the right people, helping them make sense of the options and making it easier to enquire.

That isn’t ecommerce for ecommerce’s sake. It’s simply good business.

Don’t optimise for algorithms at the expense of your brand

Being understandable to algorithms matters. Becoming indistinguishable from everyone else does not.

This was particularly relevant when the webinar discussion turned to luxury.

Jason Gonsalves of Portas argued that luxury brands face a different challenge. Their business models are built around distinction, desirability, craft and emotional connection. Being easily compared with ten similar products isn’t necessarily the goal.

The task is therefore to make the information that demonstrates a brand’s provenance, materials, expertise and story accessible to new discovery systems without reducing the brand to a list of comparable features.

There is a wider lesson here for every ecommerce brand.

AI can help customers compare. It can summarise. It can answer questions. It can make recommendations.

But brands still need to give customers a reason to want them.

That makes investment in brand, content, PR, social, physical experiences and customer experience part of the ecommerce conversation, not something separate from it.

Physical retail can drive digital demand

Laurna Woods, Group CEO at Tigerbond, sees this playing out across the relationship between physical and digital retail.

Area of expertise: Omnichannel Retail, Brand Experience & Customer Discovery

There are two forces driving retail growth at present and the brands that harness them will be the brands that thrive.

The first is the resurgence of physical retail. Our research showed brands that have opened stores in the last 12 months experienced higher brand demand growth, and popups are having an impact too as testbeds for expansion.

Yet it’s not simply their physical presence driving that growth. There is a clear halo effect that extends well beyond those four walls. Stores are marketing channels in their own right, growing customer trust, boosting awareness and brand visibility, and driving online searches and digital demand.

The second is adapting to the changing customer discovery journey. It has become fragmented, with AI playing a major role alongside social media, influencers, earned media and more. Ecommerce professionals need to get to grips with connecting their performance metrics with digital marketing, SEO, GEO, content, social, PR and broader brand building, and brand experience needs to be impeccable.

Both areas are deeply connected and together they point to the powerful competitive advantage that is community. The most successful retailers are creating meaningful experiences across every touchpoint and turning online audiences into real-world communities.

More traffic isn’t always the answer

For a long time, ecommerce growth followed a fairly straightforward equation: attract more traffic, convert some of it and keep increasing the media budget.

That becomes more difficult as discovery fragments.

Some customers will get their answer without clicking a search result. Others will research on social media, marketplaces or AI platforms before arriving at a brand’s website much further into their decision.

That makes the value of each visit increasingly important.

Samuel Shrimpton, Founder of Teylu, puts it this way.

Area of expertise: Ecommerce Growth, CRO & Customer Retention

For a decade, the ecommerce growth model was simple. Get more traffic, convert a stable share of it, bank the difference. That model is quietly breaking. Search results increasingly answer the question rather than send the click. Social platforms keep the session inside the app. The visits that do arrive are fewer, later in the decision and more expensive to win.

Most brands are still resourced for the old model. Media budget grows every year; the site stays broadly as it was built. That is now the opposite.

The implication is a reallocation, not a new tactic. If you cannot grow sessions, you have to grow what each session is worth. Conversion rate, average order value, repeat purchase. It also means owning the customer relationship through email and first-party data rather than renting attention by the click.

We saw this stress-tested properly when a restructure cut one client’s advertising budget by 75% overnight. Revenue held, because the conversion work and the lifecycle email programme were already doing the heavy lifting. Conversion improvements keep working when the media plan stops.

So at your next budget meeting, ask a harder question than “what is our return on ad spend?” Ask: if our traffic fell 30% next year, what in this plan would still be working?

Fix the foundations before adding more optimisation

Getting more value from existing traffic doesn’t automatically mean running more A/B tests.

Mathew O’Connor, Co-Founder of Actuate Digital, argues that brands first need to establish whether the ecommerce platform itself is holding conversion back.

Area of expertise: Ecommerce CRO, Website Performance & Development

Ecommerce teams keep getting sold optimisation as a testing programme. Variants, split tests, a fortnight waiting on significance. That works when the site underneath is already quick. On a lot of Magento and off the peg Shopify builds it isn’t, and all you end up measuring is how much friction you’ve built in.

Speed still gets treated as a technical concern to hand off to the dev team. It’s the first thing standing between your traffic and your revenue. We took four seconds off Helmet City’s load time during a rebuild and conversion went up 16% before anyone ran a single test. The CRO work after that added another 9%, and it landed quicker because the foundations could take it.

For 2027 planning the practical implication is about sequencing. If your frontend is the constraint, a testing budget spent on top of it is spent badly. Get a real page speed number for your PDP (Product Detail Page) and PLP (Product Listing Page), on mobile, on a normal connection. Then work out what a 10% lift on your annual revenue is worth. If those two numbers are a long way apart, you’ve found where the money is.

And ask your agency whether they’d tell you to rebuild if rebuilding sat in their P&L.

Creative testing still matters

The same principle of testing what actually works applies before a customer reaches the website.

James Bessey-Saldanha, Co-Founder at BlankTag Media, sees creative volume and testing as increasingly important to paid social performance.

Area of expertise: Paid Social & Meta Ads

Tried and failed Meta Ads for eCommerce? You’re probably not testing enough ads.

Meta Ads is a go-to for many ecommerce brands; it’s one of the best scaling channels out there and results can be fantastic. However, many brands are still struggling to make it work. Does that sound like you? Well, it’s probably because you’re not producing and testing enough creative.

Meta Ads is no longer about perfecting targeting and building a perfect funnel. Andromeda, Meta’s swanky new(ish) AI targeting system, does all of that for you. Where you’re probably going wrong is with your ad production and testing setup.

With one client (a subscription coffee brand), by doubling the number of tests they were running, within 3 months, we were able to sustainably triple their monthly media spend. This has allowed them to scale at an efficient cost and grow their market share.

This works because we don’t really know from the outset what sort of message is going to resonate with your audience on Meta, and tried and tested messages elsewhere may not always work here. Advertisers need to focus on testing multiple diverse ads to find which concepts perform best, then iterate out those winners to scale!

Use AI to create more time for the things customers value

Perhaps the most useful question isn’t how much AI an ecommerce business can use.

It’s where AI is actually useful.

During the webinar, Adam Pearce argued that brands should first identify the repetitive, administrative and data-heavy work customers don’t see. Automating some of that can free people and budget to improve the parts of the experience customers do care about.

That’s a very different approach from adding AI to every customer touchpoint simply because the technology exists.

YouGov research also shows why some restraint makes sense. While AI use is growing, trust isn’t keeping pace in every area. In early 2026, only 21% of Britons told YouGov they trusted AI in retail, while trust varied significantly according to the task AI was being asked to perform.

The opportunity is therefore not automation for automation’s sake. It is using technology to give people more time to improve the experience.

Will Yoxall, Founder of Untapped, sees that as one of the biggest opportunities ahead.

Area of expertise: Integrated Ecommerce, AI & Conversion

For years, the limit on ecommerce wasn’t ambition. It was time. Teams knew how to build richer, higher converting and more personal experiences. They just rarely had the hours.

AI changes that. It handles much of the repetitive work, from first drafts to boilerplate code to product data. The time it gives us back should go straight into building those richer, more personal experiences, because that’s still what earns the conversion.

But the real gains don’t come from any one discipline moving faster. They come when creative, conversion-led design and development meet intelligent marketing campaigns. When the experience someone lands on and the campaign that brought them there pull in the same direction, the whole thing compounds.

That’s the shift brands should be planning for. Not AI as a cost cut, but as the thing that frees your best people to join the dots between design, build and marketing.

We saw it with Stripes. Paid media, conversion-led design and development ran as one campaign rather than three separate jobs, and the account settled at over 11x ROAS, around 4 times the ecommerce average.

So, what should an ecommerce strategy focus on now?

There isn’t one new channel, platform or piece of technology that brands need to rush towards.

The bigger change is that ecommerce now sits across a much wider customer journey.

Discovery might happen through AI, search, social, PR, a creator, a marketplace and/or a physical store. Research might happen without a customer ever reaching your website. When they do arrive, they may know far more about the product and expect to find an answer quickly. And once they buy, first-party data, CRM and retention become increasingly valuable if acquiring the next customer becomes more expensive.

The brands that respond well will not necessarily be those adopting every new ecommerce technology first.

They will be the ones that understand how their customers actually discover, evaluate and buy, then work out where technology, brand, data, content, media and customer experience can remove friction or create value.

That requires a more connected ecommerce strategy.

And, increasingly, it requires different types of expertise at different points in the problem.

How we can help

At Pimento, we help brands work out what marketing expertise they actually need, then connect them with the independent agencies and specialists best suited to the challenge. That might mean ecommerce strategy, CRO, data and insight, digital and AI, brand, PR, social, customer experience or a bespoke team combining several disciplines.

If you have an ecommerce or wider marketing challenge and aren’t sure what kind of partner you need, bring us the problem. We’ll help you work out what comes next.