Future considerations of Online Marketplaces — Understanding Tendencies and Technologies. Part 2.
In our previous article, we have discussed such technologies as blockchain and virtual/augmented/mixed reality in the context of development of online marketplaces. Let’s look at some other relevant trends that have a potential to influence the industry.
AI and Machine learning
Machine learning has made significant strides recently. If 50 years ago it was a matter of science fiction and was hardly taken seriously, today it’s an integral part of our lives. From the projecting of first human neural network with electric circuits to accurate cancer diagnosing, from deep learning of natural net research to Google Brain etc. — the evolution of this technology boggles the mind. Currently, it has established its presence in the mainstream of multifarious spheres of our lives.
Machine learning is considered a subset of artificial intelligence (AI) that aims to train machines to understand their environments and synthesize the gained information thus progressively improving their performance in completing specific tasks. How can this technology be applied in the sphere of online marketing?
First, it can help P2P marketplaces to prevent fraud. Machine learning enables a system, which recognizes scammy patterns and takes necessary measures. Then, ML helps you predict what your users will like. This is achieved through enabling personalized search and recommendations. For instance, Etsy bases its search results and recommendations on the preferences of a specific user, alike customers, and shops that have similar profiles to the ones a user shopped at.
Mike Kirkwood, Founder of Eek, knows the price of such awareness, reinforced by trust:
“In my opinion, in future, marketplaces will be more connected end-to-end — where we will know a lot more about what individual people have, need, and want. And the supply chains will make it happen with a lot less human touches.With Eek, we’re getting very close to this reality. Since we work with a registry of products through the supply chain, we started doing auto-reorders and fulfillment. The key is trust. If the buyers and sellers trust us, this all works. If not, it’s just another way to try to force purchasing decisions.”
ML also allows dealing with prices effectively. Particularly, it helps to manage how to set prices for unique goods and services and how to adjust them to market changes. For example, Airbnb offers the service of lodging in the homes of almost unknown people. How can a person determine the cost of staying in one’s room for a night? The company’s Product Lead Dan Hill said that at first, this problem was a burning one for their business, causing a negative impact on it. The situation changed for the better after the company’s engineers implemented ML on their site. It was an analytical model that predicted the probability of one’s listings to be booked at particular prices within a particular time. Based on these predictions, the system gives you relevant price suggestions that boost your chances of cooperation with the guests and helps you earn as much as it is possible.
AI and Deep Learning
A broad-ranging concept of AI is well understood in the context of the limited AI applications that are able to solve one particular task. These can be exemplified by the application AlphaGo offered by Google DeepMind, which made a splash in 2016 after it beat a professional Go player Lee Sedol.
Deep learning is a sub-aggregate of machine learning that uses some methods of ML for solving real-world tasks while using neural networks that may imitate decision-making processes, intrinsic for humans. Deep learning is a very complex technology, which is difficult to implement since it requires enormous funds and data arrays. For instance, in order to “teach” a machine to recognize a human’s face or an animal’s appearance, it is necessary to adjust an immense amount of parameters and provide a machine with millions of images of the required creature.
DL can be used for various business purposes, such as text and voice search, recognition of pen-based computing, spam detection, fraud prevention, speech recognition, translation etc. Marketplaces that decide to adopt the technology will be able to use it for targeting and personalization, pricing optimization, search ranking, advancing recommendations and enhancing customer support.
Particularly, the latter function is already successfully implemented via the work of various chatbots. These virtual assistants are able to use human language to communicate with a customer, thus identifying and resolving one’s issue. They work 24/7 without lunch breaks and don’t ask to be paid a salary. Thus, they can substitute human shop assistants to some extent.
A beauty retailer Sephora offers its site visitors the opportunity to talk to a bot that will ask you some questions and find the products that will best suit your needs. Then, Amazon has its own digital assistant called Alexa; it’s a programmed smart echo speaker that can react on users’ voices and offer products from Amazon.
Okey Menakaya, Founder of MoveSavers
“I think within the next 3–5 years, marketplaces’ trends will continue to accelerate. Especially as the gig economy is expanding into almost every sector. The reach of marketplaces is endless. AI will be a game-changing event in most MPs. There are huge amounts of inefficiencies and “time-wasting” aspects, and the use of AI can solve these issues, I believe. For example, some courier marketplaces depend on customer conversion to plan their deliveries. Predicting which customer is likely to convert and how soon a customer can close a deal can be a game changer.”
The same opinion has Roy Derks, Technical Founder at SwitchBay.com:
“I expect marketplaces to overtake webshops, forcing brands to sell their products from i.e. Amazon instead of their own website. Their leading position and scalability (less fixed costs, no need for a supply chain and no “product” design) allow them to spend a higher percentage of their revenue to marketing.
From a technical perspective, I expect AI to have a great impact. The more products and users a marketplace has, the better they can become at predicting spending behavior”.
UX
With the never-ending promotion of competition in e-commerce, today’s online marketplaces need to keep pace with the provision of an excellent user experience (UX) to their clients. Some think UX is about design only. However, besides the attractive visual appearance (which is, surely, very important), a good UX implies quick interaction and loading of the pages, easy navigation, mobile friendliness etc. Almost 60% of shoppers abandon their carts and go to a competitor in case their user experience appeared to be unsatisfactory. Keep that in mind and be scrupulous about the presentation of your website, since the competition is high like never before.
Andrew McConnell, CEO of Rented.com shares his thoughts on the situation with such a competition,
“The future of marketplaces will be one of constant flux. The reason is that the rewards are simply so outsized for the space. Whether you label them marketplaces or call them platforms, these businesses tend more than any others do towards “winner takes all” principle. Given technology and globalization are also turning any industry into a globally sized one in relatively short order. This means that ambitious entrepreneurs, as well as deep pockets, will constantly pursue new and innovative businesses and models in the space. Fortunes will be made and lost, and today’s category killers could often see themselves unseated by the next upstarts who come along. It will be anything but boring, that’s for sure.”
Getting back on topic of superb UX and web design, we cannot but mention Graze — a site where one can buy healthy snacks. A visually “tasty” appearance, smooth interaction and incredibly easy and pleasing navigation make users want to come back repeatedly. Visit it to see how a good user experience may look like.
Romain de Dion, Founder of Novatopo
“Beyond chatbots, AI etc. users will be primarily looking for quality of UX, data, and vetting of supply. When they click “Buy”, they want exactly what they have in mind to happen. New techs are a necessary condition to that but not a sufficient one. Especially for verticals, which are fragmented and heterogeneous such as sports and leisure where we operate.”
Conclusion
The majority of entrepreneurs agree that the discussed technologies offer a substantial contribution to the development of online marketplaces. AI, ML and Blockchain have already been implemented in some noteworthy applications and websites. VR/AR/MR technologies appear to be less acknowledged by today’s startup founders, but we will see whether the situation changes in several years.
On that note, our grand article is coming to its end. We heard the opinions of entrepreneurs that work in various spheres of e-commerce, and now we are looking forward to hearing YOUR opinions! Do you find the abovementioned techs worthy of your attention? Which of them are you already using or planning to use?
