Everything you do online – pay bills, buy clothes, book travel, make reservations at a restaurant or an appointment with a doctor –contributes to building your online identity and adding data points to the massive amount of information on the internet.  Companies have been evolving in the way that they capitalize on this data.

Roger Ehrenberg of IA Ventures posted an interesting blog, in which he comments on the “… difference between the mere presence of data (say, an inert corpus of data accumulated from customer transactions) and its activation (putting that same data in a form that can be analyzed in real-time to provide intelligence about trends, pricing, feature attributes, etc. and classified and stored in a way that subjects itself well for future analysis).”  When looking to the most successful companies of the past decade, it is clear that they have chosen this “activation” route.

Adding Value

Many online retailers have attempted to use customer data to improve the shopping experience.  Phrases such as “You might like…” create the impression of personalization, a friendly suggestion from someone who knows you well.  It’s an attempt to provide a curated user experience where shoppers trust the site to perform a search of all products and suggest the most relevant one based on various criteria (which can be opaque or transparent depending on the site).

Amazon was one of the first to direct users to specific products, and did so based on both your behavior as well as the behavior of previous purchasers.  It offered a way to bypass the exhaustive search process and rely on the efforts of your predecessors.   It’s rumored that Gilt Groupe presents a variety of homepages, highlighting different sales for different users.  People at the fashion start-up I worked with this summer went to great lengths to prove or disprove this and found enough variation to believe that there is some customization. However, it’s unclear the extent to which this is done or the criteria used to determine which homepage a user views.

Customer information has also been used for more direct monetization through “social advertising” which capitalizes on demographic data and each internet user’s network to target a specific audience, opposed to the more traditional AdWords which relies on keyword searches by an anonymous group of individuals who may or may not be in your target demographics.  Yipit, an online aggregator on daily deal sites, analyzes purchasing behavior and trends from across hundreds of deal sites.  It packages the data into a monthly report and sells it back to deal site operators as well as to investors.  Finding value to extract from aggregated information is a different and interesting approach to the monetization of data.

Violating Privacy?

For every action there is a reaction.  In the actions relating to big data, Personal is the reaction.  Personal is a startup that gives consumers control over their data, allowing them to regain their privacy.  Additionally, this service allows consumers to benefit financially – organizations can only access your personal information with your consent, and they have to pay for it.   The company has raised $7.6M in Series A and went into beta in March 2011.

As online behavior is disseminated to more and more companies, how will individuals respond?  Will people start to take action over the availability of their personal information online?  It will be interesting to watch how the Personal user base grows or remains stagnant coming out of beta.

Personally, I believe that the availability of data has brought value to my online experiences.  I’m happy to provide my shopping history if it implies a better shopping experience next time.  While there have been minor annoyances, such as unwanted advertising and a crowded inbox, I have not experienced anything that’s inspired me to “protect” my data through a formal service like Personal.  However, it would only take one impactful abuse of my online identity to look into using Personal.

You have a distinct online identity. An identity built through your interactions with Facebook, Twitter, Google, LinkedIn, and others. This identity knows more about you than your parents do. Your parents know a few of your friends, Facebook knows them all. Your parents know your girlfriend, Google knows everything about her. Your parents might know you went to Vegas last weekend, but Twitter knows everything you did while you were there. (Thank your friends for uploading those pictures…) Your parents know what companies you’ve worked for, LinkedIn knows every colleague you interacted with while in these roles. This might be a little scary to some, but the brave see the potential in this electronic identity to open up a world completely personalized to their needs.

In today’s world, this identity lives entirely online. You open your computer and upload status updates, pictures, and information about yourself to share with friends. When the computer closes or the smartphone is turned off, you leave this identity behind and head into the physical world. But what if this personal identity could be extended into the physical world as well? Let’s think about what a day would look like in this futuristic new world.

Your morning starts with your alarm clock (synced with your Google Calendar) waking you up an hour before your first appointment. As you stumble into the bathroom to take your morning shower, a sequence of events is triggered. First, your coffee pot checks the time of your last Facebook activity from the night before and realizes you only got four hours of sleep. It automatically starts brewing a double dose of dark roast to get you through a rough day. At the same time, a touch screen monitor in the shower loads your schedule, email, and a list of important items due today while simultaneously setting the water temperature to your personal preference of 112 degrees. As you get through your shower, the thermostat, aware that you like the bathroom a toasty 80 degrees when you exit the shower, raises the temperature. Your car, realizing you’ve got to be at work in an hour, begins checking the traffic and construction reports on your route to work. It notices a detour in your route and sends you an email, which you get in the shower, warning you to leave early. As you exit the shower, your closet’s built in weather center checks the weather and learns it is supposed to be a chilly forty degrees today. It automatically rotates the clothing to bring your favorite cold weather attire to the front, keeping in mind your preference for gray suits for client meetings like the one you have today. As you zip your coat, grab your coffee, and head for the car, the smart kitchen checks its inventory versus your standard lineup and automatically places orders for bread and cereal.

Not bad for an hour’s worth of work. And the best part is that this is just a start. As your personal identity continues to grow and more devices gain access to the cloud, there is no limit to the potential customization of the world around us. Imagine every device you interact with being tied to your personal identity and fully customizable to your preferences. Every couch has adjustable lumbar specific to the user. Every TV adjusts volume, brightness, contrast, and content to the user. Windows and fans in every room automatically open and turn on to suit the particular user. Every vehicle adjusts handling characteristics to fit the specific driver. Every restaurant adjusts the menu to fit your culinary preferences. The possibilities are truly endless. So in the end, maybe it’s a good thing that your personal identity is becoming so detailed. I would certainly trust it more than my parents to pick out my clothing for the day.

For years, websites have used personalization to enhance their customers’ experiences.  Personalization has often been based on on-site behavior, including items browsed or purchased, or stated preferences.  Recently, a number of new fashion websites have begun to use personalization to identify items of interest to customers; however, I am concerned that personalization techniques used on these websites may not improve the shopping experience and may impede sales.

Although fashion websites use personalization in many ways, two types of personalized sites have become popular recently.  Both types of sites use surveys to collect customer preferences at sign-up.  These preferences impact product recommendations.  First, there are sites like Google’s Boutiques.com and ShopItToMe.com that aggregate products sold by other retailers.  When customers sign-up for these sites, they are asked to fill out a survey.  At Boutiques.com, members are asked their likes and dislikes for different styles, designers, silhouettes and colors.  Based on responses, members can view products “inspired by their style” and boutiques curated by celebrities and stylists.  At ShopItToMe.com, members are asked about their favorite brands and their sizes.  They only view products from those brands. 

The second type of personalized fashion sites use stated preferences to recommend products specifically made for the website.  For example, Shoedazzle.com members take a short survey at sign-up that asks for style and brand preferences.  Based on survey results, members review a monthly boutique which highlights product they are most likely to enjoy.  All product is designed and produced by Shoedazzle.

The primary goal of personalization on these sites is to make it easier for customers to view items they will like; however, I have concerns about their ability to identify products based on my experience in the retail industry.  My main concerns are:

-          Limited Brand/Designer Loyalty – In online shopping simulations, I have viewed how customers shop, and I have seen that designer/brand is often not part of customer shopping criteria.  Additionally, customers often have a difficult time identifying their preferred brands/designers.  This is especially problematic for ShopItToMe.com as the site only shows product that falls within the customers’ specified list.

-          Misalignment of Stated Personal Style with Actual Purchases – In focus groups, I have watched many women describe their personal style.  At the same time, I have been able to view what they are wearing and what types of products they like in magazines and advertisements.  There is a surprising amount of difference between their personal style assessment and what they actually buy.  As a result of this difference, these personalized fashion websites are likely showing products that closely align with their members’ stated style that may be very different from what members would actually purchase.   

-          Impact of Trends – Additionally, these sites are often very focused on showing trends that are most relevant for their members’ stated style.  There may be additional trends that are attractive to customers that they never see based on this focus which may limit sales.  In a real world example – faux motorcycle jackets consistently sell well at Ann Taylor.  Motorcycle jackets are not typically associated with the Ann Taylor style, but customers still want them.  If an online boutique never shows items not associated with a specific style, customers aren’t able to make these unexpected purchases.

Results from Google’s Boutiques.com indicate that the site has not held customer interest.  In January 2011, the site had 2.6 million visitors.  In April, the site only had 170,000 visitors (June 13 – Signature9.com).  And, in September, Google announced that the site would be rolled into Google Product Search (September 23 – searchenginewatch.com).  Although other factors contributed to the decline of Boutiques.com, an inability to consistently show relevant product is likely to have contributed.

Although Boutiques.com is ending, the other fashion sites have an opportunity to re-evaluate their current strategy of using up-front preferences to personalize their product assortment.  Although collecting information from customers up-front may be helpful, these sites must realize that customers cannot always tell them exactly what they want.  They need to realize this limitation in stated customer preferences and offer a wider assortment when members sign-up.  Then, they can further refine assortments based off actual browsing and purchasing on the site.