Q&A with the ALDO Group on the ROI of Advanced Analytics (Part 2)


In our latest webcast, Marc Chretien, Sr. Director of eCommerce Operations at the ALDO Group, shared his perspective on how the company approaches order fulfillment and advanced analytics in today's evolving retail landscape. 

Last week, we touched on the first part of the live Q&A, which consisted of insights around: 

  • the potential and complexity of optimizing store inventories for online order fulfillment
  • the unique approach the ALDO Group takes when fulfilling online orders from stores

In this post, we'll share the second half of the discussion to learn:

  • The process the ALDO Group took towards adopting advanced analytics
  • How ALDO Group dramatically increased sales, margins, and conversions with Celect Fulfillment Optimization 

Read on to learn what Marc has to say below: 

Q: Regarding analytics in particular, what role do advanced analytics play at ALDO Group in terms of driving success?

There are two culture changes you need to have.

One is the culture change within an organization to embrace store fulfillment. You have to have confidence in store fulfillment, and I’m here to say it is achievable and definitely beneficial. 

The second culture change I think was significant at the ALDO Group was embracing advanced analytics.

We now recognize that what was before, especially in fashion, much more heavily influenced by art there is actually science in there. There’s a science opportunity now that data availability is much more present through technology.

That’s not to say we’ve completed our journey, but I’ll say that we’re on that journey. 

We’re still going through an evolution of changing the technology across the whole supply chain spectrum, from store transactions to other digital assets, but the 'culture change' is past. Now it’s just a matter of project planning, making sure everything’s connecting properly, everything's humming together well, and making the most out of the technology. 

Q: What was it that prompted you to look at advanced analytics solutions for fulfillment? What attracted you to Celect Fulfillment Optimization?

It was really an evolution.

As we were doing store fulfillment, overcame capacity issues, and experienced the culture changes necessary to ensure store fulfillment was beneficial, the questions emerged naturally.

We started to see the “hot” seller opportunity where certain pockets of stores were fulfilling online orders with product that was popular in-store too.

After an analysis, we quickly realized all these other stores were stocked with that 'hot seller' inventory, and it was such a shame the basic OMS logic wasn’t advanced enough to consider those locations. 

Similarly, we would see this again at the end of the season when we’d go through our recall process and consolidate all the inventory leftover. There was so much wasted, leftover inventory. Wouldn’t it have been nice if we’d had that on the East Coast versus the West Coast, the North versus the South, etc...

After some hindsight analysis, we realized we needed the ability to insert decision-making in real-time or near real-time — that was the opportunity. 

Q: What was the evaluation process like? Where did you start? How long did it take?

In all fairness, Celect wasn’t the first consideration. We weren’t even aware of each other at the time.

We started with our OMS. Then we started a vetting process through industry conferences, asking questions and sharing knowledge with other retailers, industry collaborators, or even competitors out there. After some reference calls, the key appeal to Celect (and the reason I'm happy to speak positively about them today) is their system was production ready. When I say production ready, I mean it was ready for Black Friday and Cyber Monday environments. That was really important for us because if we knew we could get through peak season with this technology in place, then we would be good for 365 days out of the year. 

That’s really how it happened.

We were going through through the vetting process quite passively until we reached Celect. There was a collaborative nature to our discussions and the passion for the technology comes through from the Celect side. There was also tolerance and patience in allowing us to ask our questions to ensure any concerns were understood.

We started with a pilot program our first year and got through peak season with zero down time – which is amazing. We were able to really hone in on the order processing lag, which gave us the confidence to pursue the relationship we have now. 

Q: Where does Celect fit in your business? How does it integrate with your existing systems?

So this is a basic, simple flow (Figure 1). 

All the digital orders come in – and just to make sure everyone’s aware of the opportunity, it’s not just e-commerce. For Aldo, as an example, as I’m sure for many other players too, we have an in-store order capturing opportunity. 

(Figure 1)

Let’s say you walk into a store and we don’t have the right size, but you try on another color. Now you have confidence that you like the style, you know it fits right, and now you can place an in-store order for that color – the right color, the one you originally wanted. It’s a save-the-sale kind of opportunity, but it becomes a digital order that then gets fulfilled by this exact same network.

Whether it’s through a mobile app, e-commerce, or these in-store order opportunities – all these digital orders become eligible for this kind of optimization.

In this flow, all digital orders go through an OMS, and you are able to do some level of rule fulfillment, which is usually transportation or distance related. However, we were unable to move that to a multivariable environment where the decision can be bounced across many different types of variables in real-time. 

Considering what we just highlighted, let’s look at a slightly adjusted scenario (Figure 2) here to show you the interaction between Celect and the OMS.

(Figure 2)

The feed that goes from Aldo to Celect is data through the integration with the OMS. We’re feeding data for Celect to consume to do the analysis of what is the best decision.

The real-time integration around an OMS says: "Okay I have this order, what do you want me to do?"

Celect does the multivariable calculation in real-time and returns answers in a split second. This is true even for our Black Friday and Cyber Monday environment,  where we had about 350,000 units going through in about a four day send.

I did the quick math, and interestingly, it was about a unit a second.

It doesn’t sound quite as impressive when you break it down per second, but it is amazing when you consider every second there’s another unit and another decision. You have to those decision made in less than a second, otherwise you’re going to get a backlog. That’s a testament to Celect’s technology, not just the logic, but also the hardware.

Q: Can you share any of the successes you have had so far in terms of ROI and business impact? 

Speaking to the greater discussion around the collaboration and flexibility between us and Celect – we really wanted to understand how decisions were being made, what’s going on, and really quantify the benefits there. 

We didn’t want a "black box" technology, but rather, as the new phrase goes, we wanted a '‘glass box."

These were good discussions.

We heeded at times because the Celect team is so much more immersed in the science, but in the end we got a great place where the benefits were quantified and we understood what was there. Now we’re pushing them to introduce more variables into the algorithm and make the demand prediction even better.

The biggest benefit driver, for us, came from incremental conversion.

When the order volume is flowing through and you decide to pick store number four over store number one, I should be able to see an incremental conversion because of that decision. We set up a five week time frame and were essentially able to track the incremental transactions we had in store number 1 because we chose store number 4.

We've seen a 6 to 1 ROI, in terms of the program itself, and about a 12% incremental opportunity across our digital orders. 

Up to 12% of the time (it varies because we analyze every two weeks) we can see where we would have made a less optimal decision and probably would have lost a sale. This means we would have pulled inventory out of a store, but instead we allowed an incremental transaction.

Across our digital orders volume, we saw 12% opportunity to potentially make the wrong decision, but instead we made the right one.

This doesn’t even touch some of the other benefits too, for example, the transportation benefits. For us it was less of an indicator because our product is quite standardized where it’s all roughly the same size, but I can imagine for a different product portfolio there could be a transportation optimization factor as well.

We still get some of those benefits, because at the end of the season we're recalling less. We're consolidating less of the leftover stock. The big driver was the incremental conversion, but transportation can certainly speak to other people as well.

This interview has been edited and condensed.

*P.S. Don't forget you can always access the full on-demand webcast recording below!

New Call-to-action

Topics: ecommerce, brick-and-mortar retail, fulfillment optimization, cross-channel demand, advanced analytics, customer preference, order fulfillment, demand prediction

Ready for more?

Request a Demo