43 ways AI will impact eCommerce

Image credit: sentient.ai

The impact of artificial intelligence on eCommerce will far exceed the current impact being driven by mobile. Need proof? Here are 43 ways AI will impact eCommerce.

eCommerce is booming and the trend is projected to continue into the foreseeable future. In fact, eMarketer estimates that, “worldwide retail ecommerce sales will reach $1.915 Trillion, and double-digit growth will continue through 2020, when sales will top $4 trillion.”

Plus, take into consideration the momentum and investments being made in artificial intelligence (AI). Here are a couple of the most popular data points.

• IDC says that by 2020, worldwide spending on AI will reach $46 billion, up 768 percent from 2016.
• According to Gartner, customers will manage 85% of their relationship with the enterprise without interacting with a human, by 2020.
• A study by comScore predicts that by 2020, at least 50% of all web searches will be conducted by voice.

That’s why the intersection of these two spaces is so exciting!!

Now, I’ve read a number of posts over the last 12 months on the 3, 5, 7 ways in which AI will impact the eCommerce space in the coming years. However, I feel those numbers mis-represent the true potential. So, I took a step back to analyze the space and develop a more complete view. Outlined below are 43 ways in which AI will impact the eCommerce space. I’ve attempted to categorize them by department, and other natural segments where applicable.

Digital/eCommerce

Personalization / Hyper-personalization

1) Product Personalization

2) Engagement Personalization

3) Web Interface Personalization 

Search

4) Semantic-based Search

5) Visual Search

6) Voice Search

7) Intelligent Search Rankings

IOT (Internet of Things)

8) Smart Home – Security & Appliances

9) Wearables – Health

10) Autonomous Vehicles

Marketing

Acquisition

11) Channel Preferences Identification

12) Omni/Multi-channel Marketing Execution

13) Predictive Marketing

14) Campaign Optimization

15) Retargeting

Retention & Resurrection

16) On-boarding

17) Engagement

18) Monitoring, Alerting & Notifications

Sales

Revenue Generating

19) Add-on / Cross-sell Recommendations

20) Instant Checkout

21) Chatbots – Conversational Commerce

22) Virtual Shopper/Assistant

Tools

23) Sales process automation                             

24) Customer Relationship Management (CRM)

Business Intelligence

Analysis

25) Big Data Analysis - Understanding habits and behaviors

26) Sentiment analysis of ratings/reviews

Optimization

27) Conversion Optimization; Evolutionary Algorithms

Operations

Inventory Management

28) Smart Logistics – Restock & Replenish

29) Assortment Selection

30) Forecasting & Budgeting

Pricing

31) Competitive Price Monitoring

32) Incentive & Offer Modeling

Back Office

33) Product Identification, Categorization & Tagging

34) Infrastructure Monitoring & Alerting

35) Robotic Process Automation (RPA)

36) Decision Making

Call Center

37) Customer Service Chatbots

38) Virtual Assistants

Procurement

39) Negotiations via Intelligent Agents

Security

40) Fraud Detection

41) Identification of fake ratings/reviews

42) Counterfeit product detection

Internationalization

43) Translation & Localization

Digital/eCommerce

As you would expect, AI will greatly influence the work being performed by the Digital/eCommerce teams. The first and most impactful change will come in the form of Personalization. As you research AI for eCommerce you’ll find a lot of different terminology used for this concept including: targeting, merchandising, segmentation, discovery and filtering. However, they all basically refer to the same idea: presenting the right products to the right people at the right time.

Second, and just as powerful, are the improvements that are being made to Search functionality. For many sites, Search still garners more usage than any other navigational element; however, it’s still often a very clunky experience. On many occasions, customers leave destination sites to enter their search directly into Google along with the company name to more quickly find a page or product. With the power of AI, sites will be able to utilize semantic search to accurately interpret what you’re searching for, and will also take context into consideration. For example, a search conducted from your laptop at work will likely return different results than the same search performed on your mobile device at home. 

The other two meaningful improvements to search will come in the forms of Visual search and Voice Search. Visual Search performs a search utilizing an image that a you upload or take from the camera of your mobile device. Google Lens is terrific example of this use-case. An example of Voice Search would be Siri on the iPhone or Amazon Alexa. The goal here is to leverage Natural Language detection to understand speech, words and phrases. Finally, AI will impact eCommerce through the IOT or (Internet of Things). Examples here include leveraging AI to integrate with smart homes, connected appliances, security systems, autonomous connected vehicles like Tesla, and wearables like the Apple Watch.

Google Lens

Marketing

The second area where AI will really impact eCommerce is within Marketing. The ability to effectively identify and communicate with a customer or prospect through their preferred channel(s). And by combining the reach and frequency of those channels to break through the clutter and noise of less sophisticated campaigns. By understanding the channels customers use, their preferences for that channel, and even how they interact and move between channels. Some may say that this is too intrusive and raises privacy concerns. Yet, many contend (and this has been backed by the research below) that in general people are willing to share information and data as long as they see the benefit in the form of discounts/sales, better recommendations and improved user-experience.

“Our findings reveal that even consumers who are actively protective of their data are often happy to share it for relevant offers and value, but that value needs to be clearly defined and easily understood in order for brands to experience the greatest return on investment of various offers,” says Quint. “Young generations are much more likely to be in the Savvy and In Control group and so we expect this Mindset to grow over the next several years.” Matthew Quint, Columbia Business School - Director of the Center for Global Brand Leadership

In Growth Marketing there are three key considerations: 1) Acquisition 2) Retention and 3) Resurrection. It’s true that acquisition of new customers is critical and we’ve covered that above; however, we should not discount the importance of retention and resurrection. Utilizing AI to on-board a customer could entail utilizing notifications to remind a customer to perform a series of steps that have proven to enhance the “stickiness” of a customer. Or, to motivate their first purchase through a discount or promotion. For resurrection, it’s primarily about identifying when a customer has gone “stale,” identifying the best channel to reach them, and then providing compelling content to get them re-engaged. The cardinal rule of marketing applies…it’s more expensive to attract a new customer versus holding onto the customers that we already have today.  

Sales

From a Sales perspective there are two areas that we’ll highlight. First, are the elements that immediately generate Incremental Revenue. Amazon and their execution of cross-sell recommendations is the perfect example. Their ability to take the product the customer has already added to the cart, and then leveraging AI, make a recommendation of the next most likely purchase. All of this based upon what is known about the customer…previous purchase history, demographics, context, cookies and more.

Amazon Recommendations

Instant checkout is the concept of simply leaving a brick/mortar store with the items you’d like to purchase. The system automatically charges your account for the items you scanned while shopping. While there is certainly an offline component with presence in the store, this is still an ecommerce transaction. Prior to entering the store, a customer would establish their account, payment preferences, settings and security.  

Amazon Go 

The next two categories 1) Chatbots and 2) Virtual Shopper/Assistant have commonalities, but they also have their nuances. For instance, conversational commerce via a chatbot is most likely conducted by the customer remotely at work, home or on-the-go. In contrast, a virtual shopper/assistant would leverage the same API and services; however, it would be executed within the context of the brick/mortar store. An example here could be a virtual shopper provided by a shopping mall. It would have real-time access to products and inventory located within the physical environment of all the stores within the entire complex.

Finally, within the Tools category we have Sales Process Automation and CRM. Neither category is inherently glamorous, but each has a place in the funnel, and if executed properly, can determine the success and/or satisfaction of a purchase. For this example, think of an online mortgage company. Are you able to identify any processes/procedures that could be improved by artificial intelligence? The operations and processes of securing a mortgage from appraisal, inspection, income verification and title search all need to be sequenced and completed prior to a closing.

           

Business Intelligence / Analytics

AI in eCommerce begins and ends with Business Intelligence and data. Nothing related to AI can begin until we first identify, ingest and analyze data; and, no campaign or initiative is complete until data (results) have been measured and reported. That said, data is also one of the very first potential stumbling blocks for a successful implementation of AI in eCommerce. 

An artificial intelligence program is only as good as the data it’s provided.

No Data In = No Data Out

Garbage In = Garbage Out

Next, we’ll look at Conversion Optimization. I’ve spent a considerable amount of time working to leverage evolutionary algorithms for conversion optimization. Based on that experience, I’ll attest that the tools work. Plus, they require a fraction of the investment that would be required to execute the same tests with a team of growth marketers. Most teams have been actively utilizing A/B or multivariate tests for years, but this tool allows you to test dozens of values across multiple funnel elements to evaluate hundreds of thousands of possible winning combinations.

Conversion Optimization Data

To wrap-up business intelligence we have Omni-channel Optimization. A fully integrated campaign leveraging all potential channels (digital and in-store) to deliver 1x1 personalized content, with AI monitoring, measuring and re-optimizing in real-time 24x7. It may seem farfetched, but these are the types of solutions that are finally achievable with the recent improvements in cloud computing, big data and artificial intelligence.

This does require proper data management practices, a robust marketing platform, and executives with the vision and commitment to execute, but for the first time in our history…it’s possible.  A new business that’s properly funded will now have a meaningful advantage, because they don’t have the legacy systems and processes to slow their innovation. 

Operations

The sections above were primarily about driving revenue and growth. In contrast, but still very impactful for the bottom-line, are the ways in which AI can assist with cost control and expense reduction.

Inventory Management (Restock/Replenish) is one of the top use-cases when we discuss the topic of artificial intelligence for eCommerce. The reason is because it’s binary…do we have sufficient inventory? Now the word “sufficient” could require reams of data that would overwhelm a human, but it’s a simple task for artificial intelligence. And this same principle applies to assortment selection and forecasting.

Pricing is another application that is well suited for AI. Whether we’re referring to the development and testing of discounts, offers and incentives. Or, monitoring the prices of competitors and then developing models and revenue forecast to inform and assist the decision makers.

Monitoring and Alerting is another key consideration. In eCommerce "time is money," and on Black Friday or Cyber Monday it’s A LOT OF MONEY. So, why not use artificial intelligence to monitor, alert and even take preventative actions to stop outages before they occur. Versions of these tools exist today, but AI utilizes machine learning to monitor patterns and trends to filter out the false-positives. The result is a more reliable system, a better customer experience and fewer executive escalations.

Next, let’s discuss RPA or Robotic Process Automation. This topic exists on the fringe of AI, and you won’t see it in many other posts. I’ve included it here because it’s an area where I’ve done a fair amount of research, and have identified some potential benefits. Many articles reference people losing their jobs to AI/chatbots; however, I’d contend that the back-office workers are more likely to be replaced by RPA solutions. Here’s a 4 quadrant view of the top vendors in the space.

Artificial Intelligence for Decision Making has numerous real-world applications. In a previous role, we developed a solution that would monitor hundreds of key KPIs in real-time and “make decisions” about how and where to send the traffic to optimize our quality metric. This solution was wildly successful and generated substantial user growth for the business.

Call Center

Chatbots / Virtual Assistants have also found a place in the call center. This area has witnessed substantial innovation and growth. These solutions are heavily reliant upon natural language translation, but thanks to the Siri and Alexa the space has become rapidly more sophisticated. Basic solutions (i.e. Botsify and Chatfuel) are relatively inexpensive to build and deploy, but the development of advanced solutions leveraging Luis.ai or API.ai can become much more complex and costly. The real question then becomes…how do we determine the value? Executives are open to spending on innovation as long as there’s a meaningful return.

Chatfuel Client List

In the early days of mobile (2006) our first challenge was determining how to build a business case to fund mobile banking projects. Should we focus on sales and revenue or expense reduction? The answer then for mobile and today for chatbots is…expense reduction.

Steps to success:

1.         Identify the number one reason that customers call into your call center

2.         Develop a chatbot or virtual assistant to address those questions

3.         Deploy a call-to-action across primary points of entry

4.         Track, measure and report on the success.

Forrester estimates that the cost of a live agent call is ~ $5 each. Calculate the cost savings. How much can you save the company by eliminating just 1% of those calls?

Summary

In summary, we’ve identified nearly four dozen ways in which artificial intelligence will impact the eCommerce space. Within a decade or less AI will be so prevalent that it’ll be the foundation upon which all technology is developed. I anticipate the impact of artificial intelligence on eCommerce will dwarf the current impact being driven by mobile.

Need more proof?

AI Weekly: Google shifts from mobile-first to AI-first world

“‘An important shift from a mobile first world to an AI first world,’ declared Google CEO Sundar Pichai, summarizing the Google I/O 2017 keynote yesterday. His description of the changes underway at his company apply to nearly every business today.”

Microsoft CEO Satya Nadella on how AI will transform his company

“Microsoft CEO Satya Nadella took the stage at his company’s massive Ignite conference to lay out his vision for how deep learning and artificial intelligence will transform the company.

‘AI is at the intersection of our ambitions,’ Nadella said, noting how it will allow us ‘to reason over large amounts of data and convert that into intelligence.’ He likened AI to the arrival of books and the web and joked that we will soon create so much data that ‘we are getting to a point where we don’t even know what to name things.’”

Shining light on Facebook’s AI strategy

“Artificial intelligence has become more of a philosophy than a programming tool at Facebook. Laced across the company’s products, it holds the power to analyze data at the massive scale required by a social network connecting a quarter of the world’s population.

In a speech today at Web Summit, Facebook CTO Mike Schroepfer laid out a vision for the role artificial intelligence and machine learning will play in the company’s ambitions to improve global connectivity, technology accessibility, and human computer interaction.”

Apple's Artificial Intelligence Push at WWDC Did Not Go Unnoticed

“Apple's (AAPL) annual developer conference keynote was action-packed, with the company showing off several new pieces of hardware, including its first smart speaker. While these are likely to garner much of the attention, it was Apple's constant drum beating about its artificial intelligence prowess that may wind up being the most important.”

Are there other impacts that you believe are distinct from those above and warrant inclusion? Please feel free to share comments and feedback.