There's been a lot of talk in the media lately about the death of retail. Every week, it seems like another retailer announces the closing of stores, acquisitions or even going out of business. Many relate it to the growing competitive landscape with the convenience of online shopping and lure [...]
Omnichannel shoppers have been disrupting retailers for years, and its likely to top the industry’s agenda of challenges for years to come. But optimization, an omnichannel analytics technology, can help harness the positives of omnichannel retailing and minimize showrooming. Consider this everyday retail dilemma: E-commerce sales are growing, but in-store […]
Retail isn't an easy place to be these days. The environment is omnichannel and ever-changing. Competition is rising and retailers are struggling to understand how to best meet customers’ merchandise preferences. Fortunately, analytics are driving profitability and market share for smart retailers. Let’s take a look at the four hottest […]
Of course everyone has heard all the hype on big data and how it can help business’ become more successful. But have you thought about the different types of big data? How the different types of data can support different initiatives within your business?
Structured versus unstructured data in retail is a key topic to first understand in order to create a successful plan. Structured data is data that sits in a database, a file, or a spreadsheet. It is generally organized and formatted. In retail, this data can be point-of-sale data, inventory, product hierarchies, ect. Unstructured data does not have a specific format. It can be customer reviews, tweets, pictures, and even hashtags.
So now that you know what structured versus unstructured data in retail is, let’s talk about how to use it. Customer reviews are a great way to understand why a certain product is or isn’t working. Word clouds are a tool to visualize large amounts of customer reviews. Finding key words that are continuously being used can give insight in to product defects. For example, if ‘fits small’ is frequently used then you can be proactive by adding this to the product description or above the size selection. This will reduce customer returns and money lost on shipping fees.
Unstructured data can also be analyzed for sentiment analysis. This gives insight in to whether the customer’s response is positive, negative, or neutral. A great example of this is being able to analyze your customer’s twitter responses. Let’s say you post a tweet with products you are thinking about buying for your spring line and your brands hashtag. This enables retailers to understand your customers’ response before you even buy the product. This technique can also be used in-season and give insight to merchants on areas of opportunity or risk so that open to buy can be managed. Break down the silos between merchandising and marketing and enhance collaboration.
It doesn’t take a data scientist to use unstructured data analytical techniques either. If you’re looking to use unstructured data in your business process, check out more information on SAS Visual Analytics. Also, take a look at the 2015 Forrester Wave report where SAS was named a leader in Big Data Predictive Analytics Solutions.
Recently I read an article on National Retail Federation's "Halloween Headquarters" that 1 out of 6 millennials will dress up their animal for Halloween versus 13% of adults. With the rise in cat lovers and hipsters, I wasn’t surprised. I’m not going to lie, I once had a Pomeranian named Armani (yep, and a toy poodle named Gucci) and they both were known to be quite well-appointed furry friends.
Every year, I tried to dress my dog in a costume but he just wouldn’t have it. It was around the Paris Hilton era and all I wanted was for this little fur ball to sit in my purse and wear boots. *sigh*
How does this apply to retail? Well, does your assortment planning take into account opportunities for millennials and their costume-wearing pets?? I assure you, pet owners with those inclinations might deliver some hefty margins. And with that one factor signaling a shift among generations, it can show how knowing the age of the customers walking in to your locations and/or surfing your websites can be important for localization. Unless you are a one-location mom-and-pop shop, a historical volume based clustering approach falls flat. Only an analytics-driven localization approach can help you cluster locations based upon product selling and local demographic information. With analytics, you gain key insights into what really is driving their purchasing decisions and how best to target this audience.
Ensuring this key data element is a part of the assortment planning process is crucial to guaranteeing your millennials can find their pet costumes, and that you're diverting your pet costume inventory away from locations where they're just not going to move as well. This concept doesn’t just impact Halloween - it can have year-round benefits, and it needs to be far more sophisticated than knowing not to ship snow shovels to your Florida stores.
The analytics help you see the patterns that are not as obvious but can have a big impact, especially when you are dealing with high-margin goods and subtleties that can change from store to store. Analytics can help you provide a better customer experience for any customer at any location - that's always the goal, right?
Think about all of the different categories that differ largely by age. Dorm accessories, school supplies, clothing choices, and so much more. This will decrease missed opportunities and excess inventory as well as improving the customers’ experience. There are already 10,835 pictures on Instagram tagged #PetCostume. Where’s yours? Check out how SAS analytics can help you localize your assortments! Let's chat about your localization strategy - whenever you like. Or if you're going to the National Retail Federation's "Big Show" in New York this January we can talk live there.
Either way, I look forward to hearing from you!
Retailing holds great lessons for marketers in all industries because as long as there have been customers, retailers have focused on the customer experience. And one of the biggest elements of the customer experience is assortment - literally addressing the customer's need for variety or choice. Think of a time when you weren't exactly sure of what you were looking for - didn't you go first to places you knew had a wide variety? That's assortment.
How do retailers manage assortment? When you layer on the challenge of multiple locations, you can see how the issue quickly expands. In order to deliver a consistent customer experience, many retailers maintain a standardized layout or “floor set” among their locations - either keeping the same assortment in all locations, or varying by location. Neither approach is always better than the other, but in either case the question is how to best vary the product offerings to deliver the best customer experience.
That concept is also known as localization, and it can be quite a challenge without analytics. Early in my career, I would pull category level sales by location and put it into a spreadsheet. Then I'd sort it and make a note of the top locations by category and use that for decision making. I now know how inadequate that was. To complete the picture for you, when an order for a specific category arrived in to the distribution center, I would refer to one of the hundred sticky notes attached to my computer monitor and manually make sure that the locations received the product.
Not only was my little process painstaking, it fell short in accuracy, it was clearly inefficient, and it definitely was not scalable. With that old approach, a top volume location would likely end up on the top of the list for every category, but a lower volume store may not even get a sticky note. Sticking strictly to my sticky notes would mean ignoring the need to create a great customer experience in every location, regardless of sales volume. Quite a problem, right? But it's not insurmountable.
Using analytics, retailers are able to cluster stores based not just on demand but many other factors. Similar selling patterns of specific product attributes such as sleeve length, silhouettes, and colors can be utilized to find groups of stores that have similar customer demand. Also, location attributes such as location size or whether the store is in a shopping mall or a lifestyle center can be utilized. Even demographics of the area can be leveraged such as the average income or average age information.
Utilizing analytics enables retailers to leverage all of these factors to determine what's most relevant in the customer’s purchasing decisions. Analytics can uncover similarities that a volume based analysis may not. For example, two stores with similar volume and in the same region may need different assortments simply because of the varying tastes and preferences of the shoppers that frequent those locations. Get the assortment wrong and you miss the mark and lose sales, but if you get it right then you don't lose the sales and you deliver on a key part of the customer experience. Getting it right consistently helps you cultivate loyalty and bolster your brand. Analytics help you localize your assortment in a way that is scalable, automated, and actionable. And it will definitely save you from using a lot of sticky notes!
Assortment also affects the customer experience online, and analytics also can be applied to improve the online customer experience. Not sure where to start? Put aside your sticky notes and check out our page on analytics. Or you can take a look at the 2015 Forrester Wave report where SAS was named a leader in Big Data Predictive Analytics Solutions.
Not so long ago, I started my retail/merchandising career in the juniors division at the corporate office of a retailer. It was so exciting to be in a place where I could wear the clothes that I worked with, and I was sure that picking out cute clothes all day was what I was meant to do!
Then about a year later, I moved to the "missy" sportswear division, which I thought would be great, too, because cute clothes are easy to pick out in any area, right? Well - I thought wrong. You see, this company was based in Florida, where the missy area for retailers is long on conservative and short on trendy.
My first red flag was raised during my first “Hit or Miss” meeting, where the merchants display their top-selling items (hits) and their slow movers (misses). As you might have guessed, the very first hit was very far from a hit in my book. I just could not imagine 3,200 or so ladies actually choosing to wear a bedazzled hot pink shirt with flamingoes and seagulls swallowing up the fabric!
Sadly, one of the "miss" items was a very cute peasant top that I could see wearing myself. Suddenly I got a sickening metallic taste in my mouth as I realized how bad a fit I was for missy sportswear in Florida.
Merchandising is not as easy as one initially thinks because it involves selecting merchandise for different arrays of individuals who may not all have the same taste as you. So how do we go about figuring out what to buy? This is an age old question…
There’s competitive shopping, but relying on that always puts you behind the trend. The largest tactic for conquering this question has been analyzing what sold in the past. Many merchants try to bring items back in the next season that had mediocre performance the previous season but only to then find that they are this seasons dogs.
The customer is constantly changing and evolving. Do you buy items that you already have? No, of course not. So then how do you determine what the customer will want? That is the magical question to merchandising.
Unless you've been living under a rock, I’m sure you’ve heard the term “Big Data”. Through all of the customer transactions in store and online, there becomes a wealth of data regarding sales. Then you also have the wealth of information on social media. I think we all are still dying to know what color that dress was on Facebook. For those who might not remember, a photo portraying a dress in two different colors went viral on social media. But what if you could instead ask the audience which color they would prefer? You’d know what color of the dress to buy before you buy it! Jackpot!
Answering the magical question to merchandising is now possible through analytics. We are able to understand which attributes such as colors, patterns, fabric, or silhouettes drove your business. But most importantly, we are able to predict which combinations of these attributes will drive your business in the future, including combinations that you didn’t even have in your assortment last season! We are able to integrate social media data as well.
We are able to do this through the use of predictive analytics. Utilizing predictive analytics gives merchants the ability to predict the evolution of the trends. This truly takes the guess work out of buying! If I had these analytical insights at the start of my career, I would have been able to leverage the insights and know that rhinestones and hot pink were the next evolution of the trend!
If you want to understand how analytics can help your business and why SAS can help you, check out the 2015 Forrester Wave for Predictive Analytics where SAS is named a leader. For a broader sense of our industry-specific expertise, I encourage you to visit our Retail Industry Solutions page. Let me know what you think!
The post Predicting customer preference evolution in retail appeared first on Customer Analytics.