Skip to content
Back to Blog
Other calendar    Jun 20, 2014

IRCE 2014: Top 4 Emerging Ecommerce Technologies

IRCE 2014: Top 4 Emerging Ecommerce Technologies

2014-06-11 10.05.14

Each year, 10,000 strong attend the world’s largest ecommerce event, the Internet Retailer Conference and Exhibition (IRCE) to explore these and other online marketing challenges. This year, I was among the 10,000. Of course, the event attracts an all-star cast of speakers offering insightful talks and workshops, but I preferred spending my time simply walking the exhibition – among 600+ companies filled a 250,000 square feet at the McCormick Place in Chicago.

Following are the four most exciting themes from my experience on this year’s IRCE show floor.

1. Scalability through machine learning

A search algorithm, a data model, or a content interpreter, is only as good as the data scientists who developed them, unless the tool itself can adapt to its environment, learn, and get smart on its own. The concept of machine learning is not novel (we’ve seen applications of this concept in speech recognition, auto-correct, etc.), but its applications in e-commerce and online advertising are becoming interesting. Until recently, to gather and analyze increasingly complex customer data meant to add more data scientists to the payroll – not scalable. We are now seeing companies offering machine-learning driven solutions that enable customer data models to auto-adjust parameters when new data is collected and analyzed, eliminating the need for manual intervention. The result is faster data interpretation, lower operating cost, and higher gross margins. Watch this space, and specifically Reflektion.

2. Omni-channel predictive science – more than just the cookie diet

While many of today’s predictive technologies attempt to understand more about customers and their behavior, few really stand out. Less interesting approaches are derived from browser cookies dropped on customers’ machines. Browser cookies are good at telling from where customers were re-directed, which pages they view, and where they go after leaving your site. But with everything is shifting to mobile, browser cookies are limiting as they lack intelligent sharing across multiple form-factors (though some folks are actually working on this). Most predictive technologies today are not yet able to accumulate and interpret cross-platform data to build a true customer “profile.” However, we are starting to see some innovation in this space.

Additionally, predictive technology will begin to play an increasing role in the social space. The combination of predictive technologies and social network data (e.g., friend lists, geographical information, statuses, time of posts, likes, etc.) can be very powerful, as marketers are keen to know exactly the right time to display which content types, how often, and through which medium.

I liked what I saw from Custora and Cquotient at IRCE.

3. Natural Language Search

I am a firm believer of online experiences that promote convenience, e.g., minimizing the number of clicks required to find what you are looking for on a website. When searching through online catalogues, traditional keywords are no longer sufficient when looking for products with a precise set of characteristics - size, color, shape. Consumers tired of searching “sweaters” on a site, then filter through gender, brand, size, color, and price to find the one they want. Rather, a preferred experience is to search “Men’s medium blue Ralph Lauren sweaters on sale under $50”, which returns more relevant search results quickly. In the competition for consumer time and attention, the online retailers that can get customers to what they are looking for the fastest will have an upper hand.

The big players are already doing natural language search today. Facebook does this with its graph search today; Apple does this with Siri; Google is also doing this with their recent algorithm changes. Notice a pattern?

Natural language search for ecommerce is an important next step. With the big guys heightening consumer expectations, online retailers that do not offer it risk staying relevant and competitive.

Check out EasyAsk and Aspectiva.


4. Dynamic content and image processing

From the time a consumer receives a promotional email to the time he or she actually opens the email (if he or she opens the email) many things (e.g., product availability, discounts, customer preferences, etc.) may change, rendering contents of the email irrelevant. Today, online retailers are partnering with personalization platforms to deliver dynamic marketing and advertising content. Essentially, this means email content, among other marketing content, is refreshed at the instance of access, delivering only the most current and relevant information -- from product recommendations to discount offers. This combined with predictive sciences can be very effective 1-2 punches for converting impressions into sales. Cquotient plays here, too.

Finally, digital image processing can be used more creatively today to drive conversions. Valuable information (color, subject, content, focus, foreground, background) can be extracted from digital images and leveraged by marketers to understand the best possible experience to capture the consumer’s attention. Combine this with A/B testing and conjoint analysis and retailers have powerful way to determine the most suitable marketing experience for different customer segments. I like what LiquidPixels is doing in this regard.

These emerging technologies make for exciting times in ecommerce. I’d like to hear your feedback and comments. Please feel free to comment below or reach out to me directly at .

Kelly leads firm operations, including investment development, value creation, portfolio management, finance and marketing. She also manages investments in enterprise SaaS and fintech, serves on Edison’s investment committee, and is the pioneer of our Edison Edge value creation platform.