AI represents a major help to organizations that have to deal with large amount of data to be quickly processed. But speed does not mean quality. Check out some cases I worked on.
Retail will increasingly adopt intelligent automation technologies, according to IBM’s latest study released Tuesday at the National Retail Federation’s 2019 Big Show (). The study, surveying 1900 retail operators, focuses on the convergence of humans and artificial intelligence (A.I.) in the retail industry, and specifically how automation can help reduce human error and improve the customer experience.
IBM identified the following six ways the retail industry plans on utilizing A.I., based on respondents’ feedback:
When using Automation, Machine Learning, and Artificial intelligence I use the following matrix to define the type of effort required and identify if it is Business Process-driven or Human Insight driven. The method requires to run research and analysis about the specific organization to define the level of automation or human decision required.
While A.I. can offer great opportunities to deliver high-quality services where Customers expect on-time and reliable service (such as logistics, product availability, customer knowledge), it can produce the opposite effect when creativity and innovation are the key factors. I found this matrix useful to define the human insight need in each activity that I plan to execute. Here are some cases/examples of experiments run on Automation based on A.I. adoption:
Scoring and Propensity Modelling
Scoring and Propensity Models are the results of extensive customer data processing. I used them in the same case since in this specific case the propensity modelling was the result of the scoring method.
We developed several paths of purchase and define a propensity model based on the tracked brand points touched before getting to the ecommerce site and the landing depth.
For example, the product (details included) consistency across the touch points together with landing to the related page on the ecommerce produced a high lead scoring. A personalized flash sale with combined discount and in-store delivery based on the high lead scoring produced a high propensity.
In this case a lot of A/B testing and on-field qualitative research is involved. A.I. can help in processing data and performance but human insight is crucial to defining the propensity triggers.
MARKETING, ADVERTISING, AND CAMPAIGN MANAGEMENT
Content personalization to increase closing rate
(Multichannel Home Improvement)
The customer shopping journey of this sector is quite straightforward. It starts from owning a house and having a specific need for remodeling. The two main periods of purchase are between Q1 and Q2 and the beginning of Q4. Eighty percent of the business performance is concentrated in these 4-5 months.
Most of customer get shopping ideas on social media, home shows and showrooms. Then they go on Search engines to find reliable retailers in their area.
About 60% of conversions were no-product related. Not so many consumers have technical or interior design knowledge of products and related services, so they rely on Affiliate partners such as Home Advisor and Houzz or they trust social reviews and ask for a quote appointment based on the retailer business reputation.
40% of converted leads scheduled a quote appointment through a specific product/color page.
Here is where we used A.I. thanks to an algorithm matching the products searched on Google with the products viewed on the website. The system generated a personalized email campaign with several steps performed before and after the scheduled appointment with engagement points and further personalized promotions and contents.
The final closing was higher in this second case and made the shopping experience much quicker and easier for both the sales person and the customer.
Personalized in-store promotions
(Large consumer goods retail network)
This case was quite challenging because of the dimension of the retailer (about 3000 locations) and the low digitalization of the retail network. The personalized in-store promotions were pivotal to the success of the whole digital transformation project.
After the UX redesign of the website, the creation of a mobile app to support the newly launched loyalty program and the development of a cloud CRM system, we were finally able to push through the mobile app automatically generated personalized promotions to consumers based on their preferred store and on their closest store. The promotion content was automatically generated by an algorithm based on their CLTV, shopping list and recent purchases.
Six months after the introduction 25% of the customer base downloaded regularly the monthly coupon with a 70% actual redemption.