Consumers are no longer terrified to type in their credit card digits, but compared with legacy retail, e-commerce is still young and rarin' to go.

In the third quarter of 2017, retail e-commerce sales reached $115.3 billion, a $15.47 billion increase from 2016's third quarter. Over the course of 2017, e-commerce rose from 8 to 9 percent of overall retail -- still just a fraction of spending, but a growing one. 

The fierce competition for market share is driving experimentation, including these 2017 dot-com retail trends.

1. Independent e-commerce is thriving despite Amazon

Firstly, no e-commerce story is complete without a discussion of Amazon. The sprawling online marketplace sprawled further than ever this year, easily breaking sales records on Black Friday and Cyber Monday. Holiday spending during the end of 2017 will likely produce another win.

But two experts told Inc. that Amazon isn't everything. It's a lot, they acknowledge, but competing with the Bezos behemoth is becoming possible.

"The biggest thing people write articles about is how Amazon is sucking all the growth out of online and offline retail," said BigCommerce CEO Brent Bellm, who heads up one of the big three e-commerce software platforms for merchants.

"That story is only true at a generalized level for larger store-based retailers and multi-category online retailers. It is categorically not true for what is the most exciting thing happening in the world of e-commerce, which is lots and lots and lots and lots of smaller niche brands."

Joe Kaziukenas is on the same wavelength. He is the CEO of analysis firm Marketplace Pulse, which closely studies marketplaces like Amazon, eBay, and Etsy. "I think we will continue to see an increase in brands and retailers exploring what they can do to NOT be like Amazon," Kaziukenas wrote in an email.

Rather than waste billions trying to compete with Amazon, "the smart companies are trying to do what Amazon cannot," he added, such as investing in buzzy areas like experience, curation, exclusivity, and personalization. "A lot of this is happening thanks to Instagram." Bellm made a similar point, calling the impact of Pinterest, Facebook, and Instagram as marketing channels "colossally big."

While Amazon is known for its delivery prowess, Kaziukenas noted that even small sellers are going international. "Logistics is no longer a scary problem, but is available for everyone," he said. "Many small sellers are thus expanding in marketplaces all over the globe."

2. Consumers need to know: Is this price worth it?

Brand matters, but not always. A premium brand has to add genuine, unique value to the products that it sells, or else consumers will notice that they have cheaper options elsewhere.

RJ Jain is the CEO of Price.com, which builds tools to help consumers get the best deals across various marketplaces. "There's a huge decline of brand loyalty and the rise of generic goods," he said. "That's what happened with Wish.com or Amazon Basics."

Jain gave a nod to startups like mattress purveyor Casper and personalized fashion subscription service Stitch Fix, both of which helped push their categories forward. By contrast to the generic brands on huge marketplaces, those companies have convinced consumers that they offer the best value irrespective of the price.

3. Robots looking at fraud

The larger tech world's obsession with machine learning and A.I. has not bypassed e-commerce.

David Lennie, head of data and analytics at BigCommerce competitor Shopify, said he leverages AI-related techniques to improve Shopify's fraud-detection algorithms. Fraud is a big problem in online retail, as people use stolen payment information to buy items like expensive sneakers that are easy to resell, and merchants end up eating the cost. But if fraud filtering is too strict, it ends up generating false positives that block legitimate buyers, meaning the merchants lose sales. 

Lennie said Shopify's original fraud model was far too conservative. By training on mountains of past fraud data, "We reduced [false positives] by more than 85 percent with our first algorithm," Lennie said, and improved the identification of real fraud by 10 percent.

"It's a really good illustration of the power of machine learning and A.I. because you can build these great applications that are so scalable and a massive return on investment."

Published on: Dec 26, 2017