Empower Your People, Improve Your Retail Outcomes with Analytics
Luxottica is a company you might not know by name, but you may be wearing one of our brands on your face. We sell 93 million prescription glasses and sunglasses every year under 40 proprietary and licensed brands. We are also a global leader in optical and sun retail, with over 8,000 stores worldwide, under a dozen retail banners. So you might not have heard of Luxottica, but I bet you’ve heard of Ray-Ban, Oakley, Ralph Lauren, Chanel, LensCrafters, Pearle Vision, and Sunglass Hut.
Every company, no matter its size, wants to operate as efficiently as possible, but that becomes paramount when your company is of Luxottica’s scale. I’m Director for Retail Optimization at Luxottica’s retail North American headquarters in Cincinnati. As a company that runs retail stores in 150 countries, you just can’t be efficient without good data. That’s where my team, Merchandise Business Process Operations (MBPO), comes in.
MBPO is a small but mighty team organized into three strings: reporting and business data, market intelligence, and then the string I lead, retail optimization. Our focus is to bring advanced statistical analytics projects to business processes to make everything work better.
The Challenges Behind Our Quest for Efficiency
We faced several challenges in our drive for more efficient, optimized business operations. I group these challenges into three categories: reporting, our unique operational needs, and our desire to foster an analytics culture. I’ll address these shortly, but I’ll first start talking about the solution that helped us solve these challenges: Qlik.
I started working with Luxottica in 2012 in the EyeMed division, where I was introduced to Qlik before moving to the MBPO team in 2016. I immediately saw how we needed to apply Qlik in MBPO too. To me, it makes more sense to frame the question as: “Why wouldn’t we choose Qlik?” We had options, but ultimately went with Qlik because it hit on all of our needs.
When it comes to our reporting requirements, Qlik’s in-memory capabilities eliminate problems with constant hits to a system during peak times. Qlik also addresses our need to integrate disparate data sources and handle large volumes of data.
Many of Luxottica’s unique operational challenges related to being a global company, which means many time zones, currencies, and languages. With our centrally located Qlik enterprise installation, we can schedule reloads for times appropriate to all our global users. Qlik extensions let us toggle back and forth between languages and currencies. A centralized server node also makes it a lot easier to share all the information we need.
Fostering an analytics culture is where Qlik is a game-changer. The rich variety of visualizations available in Qlik leads to immediate data exploration and insight generation, which eases our users away from their static reports. They get excited about it, and if you can make someone excited about a change, you’re well on your way to fostering a new culture.
Visualization Answers Difficult Questions and Starts Conversations
To give you a quick overview of what we’ve built in Qlik so far, we currently have 11 apps delivered in the executive, assortment, location, and merchandise functions, and another five apps in development.
We also have a few apps for internal MBPO use, including data validation and a usage monitor. We perform daily and weekly data reloads to make sure we haven’t gone awry somewhere. We can say with confidence to our users that their data is accurate. The usage monitoring apps that come with Qlik gives you metadata about who is accessing particular pages and objects on that page, and for how long. It’s a great tool for refining apps because we can identify what users find useful, what’s not so useful, what’s troublesome, and how we can continue to improve those applications.
Let’s dive into one of the applications to talk concretely about how Qlik helps us meet those needs I listed earlier. Our Assortment Attribute Architecture planning tool (AAA) is a great example. Our assortment planners are given product attribute targets to meet. They might be told: “In this month, for this class, we want you to sell 25% of your units that are polarized.” Before Qlik, they had no way to visualize such specific targets. That was a massive problem because it meant they couldn’t understand if they were meeting those targets or not.
Qlik to the rescue. In the AAA tool, our assortment planners can easily drill down to see in this very specific data set that they sold only 7% of those units that month. The next obvious question is: “Why?”
In this same dashboard, we can look at the product assortment that we’re trying to sell, and it turns out that when we look at the three unique articles in the small-frame range, these are actually the same style—just in different colors. There’s your problem. That’s why you’re not meeting your targets.
That’s actionable information, because the assortment planners can now go to the buyers and say, “We don’t have anywhere near the assortment we need to meet this goal. We need more SKUs.” Qlik gives us the flexibility we need as developers to answer these difficult questions our users ask of the data. Simple spreadsheets just can’t do that.
That information also gives the planners some evidence to take back to leadership and argue against an unreasonable target, which is not something the planners have been able to do before.
To use another example, in Qlik we can now visualize a comparison between our sales mix and inventory mix by product attribution, to see where our mismatches are. Now the planners can say to leadership: “We need to increase inventory for this and reduce it for that.” The visualization Qlik provides opens up the planner’s interaction with leadership and informs those strategic decisions.
To return to our goal of fostering an analytics culture: giving the planners this kind of evidence to keep in their back pocket empowers them more than spreadsheets ever did.
The Final Step of Implementation Is Adoption
Qlik Sense is still pretty new to a lot of our users. One of the most important parts of my job is driving adoption, which is an integral piece of any implementation. Adoption is often overlooked in favor of all the technical things that have to come together, but it really doesn’t matter how slick your apps are if no one uses them. New users need help with new concepts, particularly when it comes to advanced analytics, because otherwise, panic and eye-glazing sets in. We use a three-pronged approach: training, app development, and evangelism.
Training is about ensuring users know how to use Qlik to its full potential. Qlik is pretty intuitive, but it’s also very powerful, and there are a lot of things you wouldn’t immediately be aware of as a new user. We wanted our users to have a good foundation as we started rolling out Qlik, so we actually customized Qlik’s beginners tutorial with our own data and examples to make it more relatable to our users.
Typically, these out-of-the-box tutorials use widely applicable sales data, which is fine for most users. But our users are specific in how they approach sales, so we wanted the tutorial to be something they could relate to. We also added sections about support, so our users would know how to open appropriate tickets at our help desk as they ran into problems.
Next in our adoption approach is app development. How do we ensure apps meet the actual user’s needs as closely as possible? First, our design team meets our users in person before we ever begin designing something. In that design session, we cover things like use, purpose, data requirements, and functionality.
It’s important to have this happen in person, because it helps spark ideas. Users might not know everything Qlik can do, so for the purpose of brainstorming ideas at these sessions, we tell users, “Assume it can do everything, and we’ll tell you when you’ve gone off the edge.”
We then set up a development stream for our users, where we can launch prototypes. This allows us to give specific access to the targeted to users who would give us the best feedback, but it also lets us eliminate any confusion about what’s in production or development.
I want to say a few things about soliciting feedback from non-technical users who don’t always know what is expected of them when you ask them to evaluate a tool. You are probably familiar with the quality of feedback when you say to someone, “Take a look at this and tell me what you think.” They leave it until the last minute and say, “Oh, it’s great.” It’s only later, when they start using it for real, that you hear all the things they hate about it.
To leapfrog to that genuine, useful feedback, I developed a step-by-step review guide. Essentially, I ask questions that the user must interact with the application to answer, a bit like a scavenger hunt. The purpose of these questions is to drive that interaction. I only ask the feedback questions after, based on their experience with the app. We’ve found the review guide gives us much richer feedback and our designers have found it an improvement to our app development process.
The final part of our adoption approach is evangelism: ensuring users and leaders are excited about Qlik. The training and app development are the technical aspects of getting users on board. I think of evangelism as being more philosophical. It can definitely represent a mindset change, particularly at a big, old company like Luxottica. I truly believe in the transformative power of a tool like Qlik, and I let that shine through.
Before we rolled out our enterprise edition to users, my team and I would frequently use our desktop editions to wrangle data, perform analysis, and deliver results. Showcasing it like that went a long way to building excitement and desire for Qlik throughout the company. That attitude is infectious. My advice is to not only be that champion, but when you find fellow champions, empower and encourage them to spread the word.
The Future Is Advanced Analytics Integration
Looking to the future, we will obviously continue developing our apps and improve that for our planning organization, but we’re also looking at ways we can expand into other parts of the business: e-Comm and marketing, for instance. We’re working on incorporating other data sources. We have some internal data sources that we will bring on board soon, but we would also love to draw in some external sources like weather data and social feeds. We could do some pretty fun things with that.
Finally, near and dear to my heart is our project of pushing forward with the advanced analytics integration. We have a whole slew of modeling projects already developed, so delivering those in Qlik is something we are excited about. We’ve already done so much, but I think we’ve just scratched the surface of what we could do.