Next growth frontier in e-commerce

Online shopping is about to change. Why is this happening, what are the implications, and how to tap into the potential? Check out CEO Toke Lund and Senior Executive Advisor in Kvadrant Consulting Carsten Pingel's take on the role og AI agents in e-commerce. 

With the emergence of AI many elements of operating an e-commerce business have been affected. However, on the consumer side and in the core shopping experience very little has changed. 

This is about to change. 

E-commerce disruption is coming in thick and fast, and it’s not brought by Amazon, as so many consultants have been preaching in non-Amazon dominated markets for years. Instead, it will be the AI agents already accountable for more than 25% of the web traffic in December 2024 alone. 

And just like in the last major growth spike in e-commerce coming from customer side experience improvements, the enabler is data. 

In the last major growth wave, mastering Google feeds and metrics throughout the customer journey was a comparable disruption. This was a massive shift for e-commerce companies, with core competencies moving from mastering SEM to mastering data feeds. Many industry professionals who remember this shift will also recall that, for a time, it allowed frontrunning companies leveraging feeds to perfection to gain extraordinary market shares in a competitive market with limited opportunities for differentiation.

Now, once again, we are standing in front of what looks like a similar moment in time for the e-commerce industry.

Why now? What is so special? And is this just one more AI promise? Yes, still a promise… but hear us out anyway. 

The role of AI now and in the (near) future

Multiple AI use cases for e-commerce companies already exist. Focus is on improving customer service, efficiency of content production etc. But typically, they are examples on more operationally efficient ways of doing what everyone has been doing for the past 10+ years. They are not exactly examples of radical AI leaps.

AI is maturing and focus is now finally coming to the application layer. Specifically, there is a high degree of likelihood that one of the first groundbreaking commercial use-cases for AI in the e-commerce market will be affecting the heart of e-commerce – i.e. search and the customer journey will change radically.

In all simplicity, online shopping is like shopping in general. So, let’s use this metaphor. For many years, Google has defined the “shopping centre” and what the “shopping window” should look like, and no one could bypass that gate. But now, we might all be starting to look at the shopping window in other interfaces and much more intelligently using Gen-AI. 

With ChatGPT’s new launch of integrated shopping, change is truly happening now. Many have described this as the Google shopping killer – but that is not the interesting part – customer behaviour will likely be stickier than expected (short-term overestimation, long-term underestimation). 

The interesting part is the opportunities provided from this – the requirements to succeed and how this could shape a new breed of future winners that take the extra leap in a competitive marked. 

The change

This next wave will also depend on structured and high performing data. But this time the core discipline will be serving the AI agents of the world – not just one feed from one company to one company (Google).

But what has changed?

•    Users can now shop directly via ChatGPT
•    The experience starts and takes shape in ChatGPT
•    This applies to both logged in and anonymous users
•    The transaction is still on the retailer's site
•    Content and recommendations based on conversation, not keywords

Just imagine serving multiple independent “concierges” (i.e. AI agents) looking to deliver in the same way as Google (as monopoly concierge) has done for years. If you are on their radar and deliver on their thinking – the “AI concierges” will propose your company and products.

The opportunity

Basically, e-commercers should optimise for a world requiring direct answers – not hitting and covering keywords and ranking highest:
•    From 100 parameters in a feed – to endless contact-points
•    From narrow search with keyword to conversation and broader context of factors being taken into consideration than ever
•    Search data driven – to both search and browsing being data-driven and in a new layer on e-commerce sites
•    Data to be content and content to be data – your content is just another data point to serve the AI agents, and in order to bring best AI experience you need really rich data and many data points.
•    Good product information is being re-defined – not just a feed – your need very rich content and product information to serve the different AI Agents, so they can serve their masters most efficiently.

Are you ready for this change?

How to start winning?

Master data in a new way and in a much broader context will be key to win. We believe the winning game will be mastering a next level of shopping feeds – such as being ready for MCP (Model Context Protocol) servers and A2A (Agent to Agent). MCP is an open standard (think “USB-C for AI integrations”) that allows AI models to connect to many different apps and data sources in a consistent way. This enables AI agents to understand and read your data better and more efficient, i.e. your way to control how LLM models understand you better.

MCP contains the world of opportunities for those who understand and manage to serve an AI agent in the future. PLEASE DO note, not building an AI agent, but serving AI agents with relevant and high performing responses, so they can serve their masters efficiently.

So, you need to start discussing a few fairly simple questions within your company: How do we service AI agents in the future? What data are they allowed to see? And how can we help them understand and navigate our data better than our competitors?

The short answer is MCP servers. The long answer is performance at scale, decoupling of backend systems, and metadata for understanding your unique data models as good as possible… (*Whispering Enterspeed*) 

Implications

Core great customer service will never go out of fashion (surprise). But we believe this will kill the many insignificant loyalty schemes out there. Because AI agents could not care less about this and 1st party data and cookies – they only want to serve their masters. So how a loyalty scheme in a world of concierge services works effectively is still to be seen. But no doubt many investments in loyalty and google feeds will be worthless in a couple of years 

And we expect massive consequences for many companies, business models, referral fees, Affiliate revenue, and consultancy fees. 

The interesting question: When?

Well, it might be here already. We believe Google and others are starting to find preferences to those companies that can deliver A2A. And this will be a factor on par with Core web vitals within 12 months.

Conclusion

Customer journeys in e-commerce are changing – your focus and competencies need to follow. It’s that simple.

Commercial excellence in e-commerce will change radically. This really is exciting times…

The rule of mastering e-commerce is still the same as yesterday, but very soon there will be a completely new prioritisation, and “Serving AI first” companies will win tomorrow’s trade. User experiences and customer journeys are “just” being redefined, but with massive implications on your GTM efforts. 

This means that mastering the basics in terms of price and logistics are still the shoulders you stand on – and mastering performance marketing and loyalty in the old paradigm we have lived in for at least the last 5-10 years will not be enough. You need to start thinking about serving AI Agents – or give us a call 😃

 

About the Authors:

Carsten Pingel has been working within digital strategy and business development for more than 15 years. Former experience includes acquiring and scaling high-growth, multi-brand e-commerce businesses, and enabling digital transformation in the largest Nordic media group, Egmont, business development and strategy at Copenhagen Airports, head of strategy consulting in Valtech, and board positions in high growth e-commerce and SAAS companies. Now Carsten spends his time as Senior Executive Advisor in Kvadrant Consulting

Toke Lund has been working within management consulting, heading up the strategy office and the consumer insights office at the LEGO Group, driving digital transformation at Salling Group, and working as a digital strategy consultant at Novicell. Now Toke has founded Enterspeed, a software company delivering a high-performance product called Speed layer, which can accelerate digital transformation and composable setups. 

Together, they've also written the article Surviving the Ecommerce Industry – Porter style and the article Profit before Growth about massive digital business leadership challenges.