Generative AI (GenAI) has enjoyed quite the hype cycle, and businesses – around the world – are considering what they can do to equip employees with safe GenAI tools.
ALDO Group has these conversations on a day-to-day basis. Yet, GenAI is not at the top of its priority list.
“At first, I was very excited,” recalled Matthieu Houle, CIO of ALDO Group. “It’s a new technology; what can we do with GenAI?
“It reminded me of the mobile revolution 15 years ago, and we considered the cowboy approach to it. But, pretty quickly, we had to come to terms with how to tackle this whole thing.”
Soon, Houle realized that GenAI wasn’t something his team would have to chase. Instead, it’s something that would come to them.
“Yes, it was super exciting for me and the team to work on possibilities like building a chatbot for employee service,” he continued. “But, companies like SAP will bring this to our people, as they already have with other GenAI use cases across different functions.”
My role is to create differentiators… So, with GenAI, yes, it will bring new productivity and help, but most people will have access to the same tools, in general.
As such, instead of chasing the big GenAI dream, Houle understood that his time was better spent elsewhere, on good, old-fashioned predictive AI and machine learning.
“The more transformational AI for us is more on the predictive side,” Houle added.
For instance, we have a project around predicting the future of sales. That’s the “Holy Grail” of retail. You can do a lot of things, such as optimizing your inventory, price, channels, and all of this stuff.
The project Houle mentioned culminated in the launch of Delphine: a predictive machine learning model that’s supporting front- and back-office teams across ALDO Group.
Meet Delphine. ALDO Group’s Big Predictive AI Investment
Delphine is the scientific brain behind ALDO Group’s demand forecasting.
For ALDO Group, that engine is critical. Just think of the nature of fashion. The customer thinks: I want a product at the right price, which makes me feel good. That much hasn’t changed.
But, what is changing is how consumers get inspired, decide to buy, and where they want to buy.
So, from a brand and reach perspective, the critical question is: how can I better predict demand so I have the inventory in the right place, in the right size, and at the right price?
Recognizing this, Houle stated:
To be able to predict demand and use predictive AI in the back office, I think – other than huge financial savings, less of a carbon footprint, and less of a cost – that’s the biggest lever we can pull to improve consumer experience and differentiation.
Delphine will prove crucial here, but not as an all-knowing supercomputer. Instead, ALDO Group sees her as a colleague – hence the name – who has a seat at the table, helping the team crunch data. As Houle said:
Delphine won’t replace the buyers, the planners, and all these people; you need to bring her to the employees to help them do their work.
Houle recognizes composability as a critical trend that enables this predictive employee support. He needs to embed Delphine into employee workflows and interfaces to get the team to participate.
As such, he notes SAP’s switch from a more monomythical planning application to a composable blueprint that his team can adapt with its data – wherever it’s coming from – as a big time saver.
Where ALDO Group Believes GenAI Can Add Value
While Houle trusts ALDO Group’s tech partners to bring new GenAI use cases to them, he understands the technology’s potential to support employees and drive productivity.
“If you don’t do it, you will become a laggard,” he acknowledged. “Like, if you didn’t have an eCommerce website in 2020, you’re probably not in a good position today.”
As such, ALDO Group has implemented several use cases, including auto-generating product descriptions for human review.
However, Houle’s team also developed a more innovative solution after the company ran a hackathon at the Université de Montréal.
At the event, the company asked students to build retail solutions using GenAI. Houle explained:
The team that won built a product where a consumer could say: “I want shoes like David Beckham would wear,” for example. It then takes this style-based approach to generate an image and recommend options for shoes in a conversational manner.
Now, ALDO Group is bringing the product into production and bolstering its eCommerce proposition.
Deploying Predictive and Generative AI In a Responsible, Non-Creepy Way
With tightening data privacy regulations and global bodies – like UNESCO – releasing ethics guidelines, responsible AI is a topic that’s continuously hitting the headlines.
CIOs like Houle are acutely aware of these stories and the risks of AI. To avoid those risks, he follows several rules. The first is to tell customers precisely what data ALDO Group is collecting and what it does with it. Houle stated:
What’s important is to be upfront with customers about what’s going on, but also give the ability to the consumer to say that something was wrong here.
Interestingly, the CIO noted that GenAI may help here, offering brands an opportunity to be more direct with customers and avoid coming across as creepy.
For ALDO Group, that involves transitioning from implied personalization – aka. trying to figure that the customer has clicked on this, so they must like that – to specifically ask the customer: do you like this shoe?
“That’s a massive trend, largely because of privacy and customers realizing that there was a lot of creepy stuff going on within the industry,” said Houle.
There’s more of an acceptance towards more implicit data sharing and taking a bit of time to answer questions at the right moment via a chatbot.
In making this statement, Houle likened this implicit conversational feedback experience to following a truck on the highway with a “rate my driving” bumper sticker.
So, if the customer senses that the AI application is going wayward, they can “rate my AI” and help the business improve its steering with more sophisticated guardrails.
He concluded: “Instead of trying to hide and believing this is going to be perfect, accept that no, it won’t. But, guess what, that’s okay.”
Bringing AI to the Workforce: Best Practices
To finish up, Houle acknowledged the real concern of employees across many organizations that – one day – AI will take their jobs.
As such, there is a risk of employee reluctance to leverage the AI that IT teams have injected across their desktop and workflows.
To help brands avoid this, the CIO shared a few best practices, with the first to position AI as a superpower, not a replacement, and involve them early in the thinking process.
“When we have an early AI project, the end-user – who is going to benefit from that superpower – is at the table with us,” Houle noted.
Next, with those employees, Houle suggests going through their to-do lists and considering: if I took away this, what else could you have time to do that adds value?
That’s an excellent way to empower employees and inspire them to get excited about having those AI-fuelled superpowers.
Finally, the CIO noted that it’s all about training and change management, so when IT builds something, it’s not rejected by the team.
Thanks to SAP for arranging the virtual session with Matthieu Houle and ALDO Group.
For more insightful interviews with customer experience and technology leaders, read our articles:
- Swisscom Snaps Agent Workloads In Half After Running These Two Projects
- Holland & Barrett’s Director of CX On Customer Satisfaction and The Importance of Market Research
- “Eat Your Own Dog Food!” – Lessons from a Chief Experience Officer