Why AI for Digital Commerce still needs a human touch

Orla_Power_Profile_PictureBy Orla Power, Director of Marketing, Luzern eCommerce

With eCommerce continuing to accelerate, brands and retailers are digitising faster than ever before.

Most eCommerce brands are implementing, or have already implemented, digital transformation strategies in order to remain profitable in the face of fierce and fast-changing online competition. The integration of artificial intelligence (AI), machine learning (ML), and algorithms into their systems is becoming increasingly important to support this change. According to Gartner, worldwide IT spend is projected to total $4.2 trillion this year as organisations reshape their operations to satisfy the demands of the new landscape. 

Many companies, however, are playing catch-up, trying to manage the vast complexities of maximising their online sales revenues, whilst also future-proofing their systems to cope with the pace of change.


Real-time shifts in pricing, product site rankings, and consumer behaviours have become critical factors in an online brand’s viability and profitability. However, the measurements, analysis, and resultant actions necessary to stay competitive have become both complex and time-critical to manage without significant algorithmic assistance. Yet the sophistication required with algorithms, as well as the human experts necessary to maximise their usefulness is beyond the resources of many brands’ in-house teams.

AI & eCommerce

In our recent survey of eCommerce decision makers, a majority (52%) of eCommerce decision-makers surveyed see AI tools that automatically take action on eCommerce market and shopper data insights as an opportunity to improve their eCommerce strategy. Over half (51%) view improved personalisation (based on shopper behaviour, previous actions, purchase history, and other enriched data) as an opportunity to improve their eCommerce strategy. What’s more, over 2 in 5 (46%) of eCommerce decision-makers surveyed agree that analytics around the competitive seller and category analysis is also an opportunity to improve their eCommerce strategy.

According to leading management consulting firm McKinsey & Company, the “high-performing companies” that gain added value from AI are expected to continue to invest in technology post-pandemic. A whopping three quarters (75%) of eCommerce business decision-makers that we surveyed, expect their technology budget to increase to support their eCommerce strategy.

Consumer expectations

As brands and retailers continue to redesign their eCommerce operations in the wake of the challenges posed by the pandemic, it’s clear that consumers’ expectations have increased even more. Smooth and seamless online shopping experiences, backed by fast delivery and trouble-free after-sales, are here to stay. Not only is this level of service the expected norm, but today’s online shoppers have a healthy appetite for continuing improvements. This means that brands and retailers must be more customer-centric than ever before and truly become “CX obsessed” in order to meet consumer expectations.

Taking a human-centered approach to eCommerce challenges means looking at humans first, at real people and their experiences, and deriving actionable data to create better customer journeys.

AI vs Humans

It is perhaps ironic that as AI has become smarter, it has also become increasingly reliant on humans to harness its power, tailor its use, understand its findings, and manage its resultant actions. Most substantial online traders are now using AI, however the algorithms themselves are in competition and are constantly evolving, not just to keep up with, but to anticipate changes in the marketplace and the needs of the customer base.


The use of ML facilitates this unprecedented rate of change. It’s the core technology of AI which allows algorithms to learn, to predict market movements and consumer decision-making and adapt accordingly. We are not yet under threat of world takeover from AI – thankfully the self-aware machines in the Terminator films remain the stuff of fiction – but there are still very real risks attached to leaving self-learning algorithms to run unsupervised through the eCommerce landscape.

In his 2011 TED Talk ‘How algorithms shape our world’, entrepreneur and tech expert Kevin Slavin recounts how a book about houseflies was briefly on sale on Amazon priced at $1.7 million. Purchasing at this price would, he said, have been a bargain given that a few hours later its price had risen to $26.3 million (plus shipping). Behind the humour, Slavin is making a serious point about what happens when algorithms, in conflict in this example, are allowed to act without any human control over, or even knowledge of, what is actually happening.

In one sense AI and ML have come a long way in the decade since Slavin’s talk – but then in parallel the need for understanding and managing them has become even more important to brands whose online offerings, and therefore their profitability, are largely in algorithmic hands.

The complexities of successful online selling have become vast, and dashboard-level analytics are no longer fit for purpose. A product needs to remain visible online. Its sales copy, photography, and promotional videos need constant testing for optimum effectiveness. Its pricing needs to be adjusted to react to changes in competitors’ offerings and to the demands of the consumer. CRM and ERP systems need data-mining for sales trends and behavioural patterns. Reacting to these critical fluctuations no longer happens weekly, daily, or even hourly, but in real time. This is why an increasing number of brands are partnering with specialists in the field such as Luzern.

Managing these processes today requires a combination of complex, advanced, and constantly evolving algorithms and teams of highly specialised analysts, marketers, and eCommerce experts to turn them into meaningful insights that drive an eCommerce brand’s channel strategy and activation plan. To succeed in today’s digital landscape, neither the algorithms nor the humans are sufficiently effective without the other.

How does Luzern fit in? 

Luzern specialise in offering both sides of this equation; the highly advanced, bespoke AI needed to ensure brand visibility, profitability, and sustainability, complemented by the specialist human expertise and knowledge to maximise their impact. This combination, in today’s eCommerce world, is a major factor in a brand not only surviving but thriving.

If you’d like to speak with one of our marketplace experts please feel free to get in touch at hello@luzern.co.


So What’s next?

Are You Ready? LET’S TALK