Part Three: Q&A with Aéropostale’s SVP of Planning & Allocation on AI


"AI, which is forecast to grow to $36.8 billion by 2025, could bring a new way of transformation to retail." - Deborah Weinswig, CEO of Coresight Research tweet-image  

Everyday you make really difficult inventory decisions. 

The impact of AI on retail is clearly a major topic of discussion for many merchandisers on the hunt for new tech to help make their job of managing inventory and forecasting demand better, faster and much more accurate than ever before. 

At the end of the day, your goals revolve around improving revenue — and yet, so much of it is tied up in unsold product, inefficient processes, or bad merchandising decisions. 

If a technology existed with the sole purpose of helping you meet these goals, why wouldn't you use it? Many retailers are in a dire position because of this failure to adapt and evolve with the tools available to meet the customer-centric demand driving the market today.

And, at the moment, machine learning and advanced analytics are the tools making it possible to do just that. Retailers like Aéropostale are taking steps to make better investments in the right products with their customer in mind by incorporating advanced analytics into their merchandising process.

Check out what Karen Walter, Aéropostale's SVP of Planning & Allocation, has to say in our latest webcast Q&A (transcript below!) about how and why her team adopted analytics in today's retail environment.

(Also, feel free to access Part 1 and Part 2 of the Q&A transcript for more insights from Aéro!)

Q: How have advanced analytics impacted your role and the change retailers are experiencing?

It’s actually impacted it quite a bit. There was a little bit of an evolution in that impact as well.

A few years ago, this was actually a scary concept because the expectation was perhaps “I need to be an expert in advanced analytics” or “I need to understand how this data is coming together.”

How do I do that when I don’t?


I don’t have that background, that’s not what I do. How can I lead a team people if the expectation is that they are going to be experts in the actual science of advanced analytics? 

Thankfully, it evolved fairly quickly to a point where I don’t need to be an expert in the science of advanced analytics. 

I don’t actually have to hire people who are experts.

I do need to find a technology partner (who I trust) to do the advanced analytics. Additionally, we must be able to have a conversation about how to apply it to the opportunities and the business I’m trying to run. 

For my role in particular, I’m at a much more comfortable place because I can continue to lead my team and hire people with skills who have always been successful in this role. 

At the same time, there’s an understanding that advanced analytics is something we want to use, rely on, and continue to work with.

However, now I have a business partner who’s going to speak my language and the advanced analytics language—I don’t actually have to be that bridge. This brings more of the leadership and strategy to the forefront of my role, rather than building expertise in an area I didn’t have.

Q: Can you share any of the successes you have had so far?

For our fall buying season, we have our men’s and women’s apparel team using Celect to place their buys. For our holiday season, we have our planning organization using Celect to help build the plans in the first place.

That’s kind of where we are. 

We don’t have any product in store yet where we’ve used Celect as a recommender. That’s still to come. But the success so far, the key success, has been the adoption throughout the merchandising and planning organization—those are the two groups most impacted right now from this (other the IT organization that got it all in place). 

We had several town hall meetings where we shared the look and feel of the system with other groups who work closely with planning and merchandising but aren’t directly impacted by Celect. We’ve also started to introduce it to other areas of the company that can also benefit from advanced analytics (but weren’t necessarily part of the initial push to buy less inventory).

We’ve had a lot of success getting visibility to what Celect can do a little bit more broadly in the organization, and the adoption has been really strong.

I don’t know that new technology adoption is something we’d say we’re experts at. In the past, Aéropostale had a project not go very well, but this (so far!) has gone very well. I think having buy-in from the beginning, the length of time we took to choose the partner we chose and staying very focused was key.

Know what you’re trying to solve.


Any software can do a number of things, especially once you start talking about advanced analytics and data. Of course, it can be applied in so many different ways. However, we always tried to stay focused on that initial question.

We’ve started to circulate it to some other groups to get more people involved, but the teams using this system right now have stayed very true to the initial question, which was key to the success of the roll-out.

You have to do your research. You have to find the right partner and make sure you’re going to be able to work with them as part of your team. 

Q: What were a couple of your biggest struggles before even getting to the point where you could do advanced analytics?

Two main ones come to mind as the biggest struggle to even get to this point:

First, just the general idea the technology and analytics existed. Something out there could better predict customer behavior than you can with all the knowledge you have sitting here.

There’s still a little bit of the unknown that these things can work. That’s a big hurdle.

You have to be able to get the buy-in outside the simple desire to be able to do your job better but understanding that this actually works. There is intelligence you’re not utilizing, and that’s a disadvantage. 


Additionally, this will require many conversations, a ton of proving it out, dragging people to different forums, reviewing business cases. You’ll have to do a lot of legwork to just get the people you need buy-in from to simply embrace the idea there’s potentially a way to be smarter about the data you have and how it can help.

The second biggest hurdle is the cost.

These things cost money. You have to be able to figure out how to invest in this technology and prove the benefit. How are you going to pay the company back for the investment?

Those were the two biggest hurdles.

*Access the full on-demand webcast replay below!

New Call-to-action

Topics: allocation, predictive analytics, merchandise planning, inventory optimization, data, advanced analytics

Ready for more?

Request a Demo