How Claire's Is Using Humans To Ensure The Quality Of Their Data

We had a discussion with Sean MacCarthy, Global Head of Analytics and Store Segmentation at Claire's about how they are using big data and analytics to increase the company's efficiency and customer satisfaction

15May

Modern big data analysis has injected new life into retail over the last decade. New techniques are enabling stores to analyze customer data in a multitude of ways, offering personalization opportunities which would have been impossible only a few years ago, and the technology is constantly advancing.

This means the impetus has fallen on each retailer to determine how they will best make use of all this data while also respecting their customers. They need to weigh up the limitless potential with ethical and security concerns. The beauty with all of this data being is that there are so many different, potential uses - but this is a struggle too. There isn't a guidebook every retailer can follow that will help them reach their analytics goals. However, as the industry continues to grow and evolve, individual retailers are all contributing towards a collective pool of knowledge of processes and ideas that have and haven't worked.

Claire's is one such company. Like many of their competitors, they have begun to lean heavily towards incorporating data and analytics to both better their customer's experience and improve their supply chain efficiency. To explore this further, In the lead up to the Big Data & Analytics for Retail Summit, I spoke with one of the speakers, Sean MacCarthy, Global Head of Analytics and Store Segmentation at Claire's. He is the strategy and insights leader driving profitable solutions for organizations across industries through analytics/data modeling, S&OP, lean/six sigma tools and change management.

How important is data analytics to confronting the threats posed by competitors?

I like to think that roughly 90-95% of business problems can be solved with Algebra, and if your basic reporting and analytics suites for your operational roles don't give the prescriptive insights that really could be automated, then you've already failed at what are 'table stakes' for the best retailers out there. The remaining 5-10% of problems are where it'll be the data science teams and their models and how well they can integrate with existing systems that'll really differentiate retailers in the future.

How is big data improving Claire's supply chains?

Supply chains have always revolved around the biggest data in an organization. What the compute power really allows us to do is start taking into consideration all the data which were previously unanalyzable. We were limited by the difficulty in just trying to compute location/product specific insights that could then be re-grouped back into executional forecasts, allocations, etc... This is where there's a tremendous boon for companies that take advantage. Forecasts that are accurate at remarkably discrete levels and able to react quickly to trend changes that might otherwise go unnoticed by even the most experienced inventory management teams.

How is the IoT revolutionizing how Claire's operates? Are organizations prepared to exploit the increasing amount of streaming data that it is being produced?

I think the most immediate impact is currently felt on the operations side where sensor data can alert employees to issues with equipment (e.g. freezers in a grocery chain, engines in a transportation fleet, etc...). Likewise, it can help redeploy labor based on consumer demand (e.g. stock outs based on real-time feedback from RFID or video). There are really innovative things happening on the customer-facing side as well, but the ROI on these are not as widely felt yet.

How do you ensure that all the data you collect is accurate and identifiable?

This is the biggest challenge in retail analytics in my mind: quality data and knowing when it's not. We employ the typical ETL checks, but this is where the human side can play an incredible role. This is because it's often the people who are in tune with the business that will notice that, even though the values may be within a checksum tolerance, there's something else going on. This will either lead to a process to alert when these conditions are met going forward or the identification of a 'sister' metric that needs to be incorporated into a model going forward.

Have you started incorporating more automatization to ensure you are drawing the right insights promptly? Do you foresee this changing the work of the data scientist?

Absolutely, but given the retail environment, recurring checkpoints need to be put in place to ensure that any automated insights remain relevant to the business.

How important is customer trust when it comes to your data efforts and how do you earn it?

This is obviously at the forefront of many consumers, politicians and other regulatory bodies minds and for good reason. As far as the security of customer data is concerned, that's a given, if you're not doing everything reasonable with the data you have to ensure its protection, you might want to stop gathering it. But that's behind the scenes and often the trust isn't lost until something goes wrong.

To proactively gain trust, we have to understand that customers want value for what they give you, whether it's their money or their data. Just as they expect us as companies to fight hard for their money by providing unique products that bring joy or solve problems. They want us to ensure we're protecting their data and using it in ways that benefit them, not just ourselves. When those two circles overlap, they'll be grateful for the incredible, personalized experience. However, when its just for the company's benefit, then you have a problem. Hence, make it transparent how the data you collect about them helps you deliver the products, offers, experiences and, deals they enjoy.

In what ways is Claire's progressing towards a more omnichannel experience and what challenges does this present to data analytics?

I think for many retailers it's just getting the basics down at this point. Click & Collect was invented over a decade ago, but is just becoming mainstream for many retailers. Things like that, visibility to in-store inventories, etc... this is where many need to start and get right, but quickly! As for the analytics challenges, it's going to come back to having good data in the near-real-time that most omnichannel experiences require. Once you've got that, it's just securing the right methods to disseminate the offers, information or experience to the customer, be it mobile, concierge devices in store, etc...

What can the audience expect to take away from your presentation?

I believe they'll take away a fresh look on how to target customers at the moment of expression of need as opposed to the now, often too late, zero moment of truth. 


For similar insights into how retail is leveraging big data and analytics to pinpoint opportunities for business growth, attend our Big Data & Analytics for Retail Summit happening in Chicago on the 6-7 of June 2018.

Dataviz

Read next:

Why We Need Data Visualization To Understand Unstructured Data

i