For years, ecommerce strategy was relatively easy to divide into channels.
There was the D2C website, paid search, social, marketplaces and, increasingly, retail media. Each had its own objectives, budgets and measures of success, and in many businesses, its own team responsible for delivering them.
The problem is that customers nowadays don’t behave in such a neat and predictable way.
A product might first appear in someone’s social feed, be researched through search or an AI platform, discussed on Reddit, compared on a retailer’s website and eventually purchased through Amazon. The place where discovery happens is no longer necessarily the place where the transaction takes place.
As those journeys become more fragmented, the way brands think about commerce may need to change with them.
In a recent episode of The FODcast, Megan Conahan, EVP at Direct Agents, joined us to discuss the evolution of ecommerce and retail media, changing consumer behaviour and what increasingly distributed customer journeys mean for the brands trying to reach them.
Commerce is becoming increasingly platform agnostic
One of the clearest themes from our conversation was just how much the path to purchase has changed.
Consumers have more places than ever to discover products. Search, social media, creators, marketplaces, retail media networks, communities and now conversational platforms can all influence what someone eventually chooses to buy.
But discovery and purchase don’t necessarily happen in the same place.
Megan describes this as “platform agnostic discovery”. Rather than attempting to dictate where a customer should discover or buy a product, the focus shifts towards understanding where that particular audience spends time, where it is genuinely influenced and where it ultimately prefers to transact.
That distinction between attention and influence is particularly important.
A customer may spend considerable time scrolling through one platform but turn to another when they actually want advice, validation or recommendations. They might then move somewhere else again when they are ready to buy.
For commerce teams, that makes the customer journey much harder to contain within a conventional funnel. It is becoming a network of different interactions, with each potentially playing a different role in the eventual purchase.
Retail media is becoming part of a much bigger ecosystem
Retail media has become a significant part of this shift.
The combination of customer data, advertising inventory and the ability to connect activity with a transaction has made retail media particularly attractive to brands. But its role is also expanding beyond advertising to customers already shopping on a particular retailer’s platform.
Retail media networks are increasingly connecting with the wider web, giving brands more opportunities to use retailer data and reach customers elsewhere.
At the same time, marketplaces and retailer platforms have become established destinations for the transaction itself. Even when a brand has invested in its own D2C experience, customers may still prefer the familiarity and convenience of buying through Amazon or another retailer.
That doesn’t make D2C irrelevant. Brand websites can still play an important role in discovery, content, loyalty and building a direct customer relationship. But it does make it harder to assume that the ultimate objective of every marketing interaction should be to drive someone back to the brand’s own website.
The more useful question becomes: where does the customer actually want to buy?
Why channel-level measurement can give an incomplete picture
This more fragmented journey creates an obvious challenge: measurement.
When customers move between multiple platforms before buying, attributing a sale to one channel becomes increasingly difficult. Yet many marketing teams are still measured according to the performance of individual channels.
That can create some misleading incentives.
A campaign targeting people who already know a brand may deliver an impressive return on advertising spend, for example, while doing relatively little to generate new demand. Conversely, activity that introduces entirely new customers may look less efficient at the point of their first purchase but ultimately generate much greater value over time.
This is why the conversation around marketing effectiveness is broadening beyond ROAS alone.
Megan talked about looking at measures including total advertising cost of sales, new-to-brand customers, lifetime value and overall sales performance to understand what marketing activity is actually contributing to the business.
That becomes particularly important as profitability comes under greater scrutiny. Rising costs mean brands need to understand not simply whether an individual campaign generated a return, but whether the overall combination of investment is contributing to profitable growth.
The question therefore shifts from “Did this channel perform?” towards “What role did this activity play in achieving the wider commercial objective?”
AI is adding another layer to product discovery
The arrival of conversational AI makes an already fragmented journey even more interesting.
Consumers can now ask questions, compare options and research products through AI platforms before ever reaching a traditional search result, retailer or brand website.
That changes what being discoverable means.
During our conversation, Megan gave the example of product copy built around evocative marketing language. That language might resonate with a person, but an AI system also needs clear information about what the product actually is, its attributes and whether it is relevant to the question being asked.
Brands therefore need to consider how product information works for two audiences at once: the human customer and the technology increasingly helping that customer make a decision.
Consistency becomes important too.
Product information doesn’t exist only on a brand’s website. It can appear across marketplaces, retailers, reviews, social platforms, community conversations and numerous other sources that AI systems may use to understand a product.
Optimising one product page in isolation may therefore only solve a small part of the problem.
For commerce teams, this isn’t necessarily about abandoning existing search or content strategies in favour of the latest technology. It is about recognising that AI has become another layer within an already distributed discovery journey and making sure the foundations are strong enough to support it.
Breaking down the internal silos
Perhaps the biggest challenge created by all of this isn’t technological at all.
If customers move freely between channels but the teams responsible for those channels operate independently, there is an obvious disconnect.
A D2C team may be incentivised to drive customers towards the brand website. An Amazon team may be focused on growing marketplace revenue. Paid media might have its own performance targets, while retail media, content and organic search sit elsewhere again.
Each team can be successful according to its own metrics without necessarily producing the best outcome for the wider business.
Megan’s closing advice was to look carefully at those structures and objectives. Rather than asking individual teams to maximise sales within their own channel, businesses need greater alignment around the overall commercial goal.
That sounds straightforward, but changing established structures, budgets and responsibilities rarely is.
It requires teams to share information, understand how different channels influence one another and become comfortable with the fact that activity they own may ultimately result in a transaction somewhere else.
As customer journeys become less linear, internal structures may have to become less rigid too.
The capability challenge behind distributed commerce
That change also has implications for the people businesses need.
Greater automation means some of the highly manual work traditionally associated with digital marketing can increasingly be supported by technology. Megan shared how Direct Agents is already using AI to support areas including bidding, budgeting, testing, performance monitoring and creative production.
But rather than removing the need for people, that is changing where their value sits.
Megan described a growing emphasis on strategic thinkers who can understand the bigger picture, work across disciplines and translate increasingly complex information into better decisions.
That feels particularly relevant in a distributed commerce environment.
Knowing one platform inside out remains valuable, but businesses also need people who understand how that platform connects with everything around it. They need to recognise when a strong channel metric doesn’t necessarily represent strong business performance, understand how customer behaviour is changing and collaborate across traditional functional boundaries.
The technology may be getting better at pulling the levers. Deciding which levers matter and why remains a much broader challenge.
Final thoughts
Commerce isn’t moving from one dominant channel to another. If anything, the opposite is happening.
Customers have more ways to discover, research, compare and buy products, and the boundaries between those stages are becoming increasingly difficult to define.
That makes it harder for brands to control the journey, but perhaps control is the wrong objective in the first place.
The opportunity is to understand how customers actually behave, make it easy for them to move between different touchpoints and ensure that the business is measuring success against the overall commercial outcome rather than the performance of individual channels in isolation.
AI, retail media and emerging platforms will continue to change the mechanics of that journey. The bigger challenge for commerce leaders may be ensuring that their strategies, teams and measures of success evolve with it.
A big thank you to Megan Conahan for joining us on The FODcast and sharing her perspective on retail media, platform-agnostic discovery, AI and the changing shape of commerce.
Listen to the full episode below:
