In December 2021 McKinsey published a report entitled “The state of AI in 2021” on the use of AI by global players.
56% of respondents indicated that they had implemented at least one AI-supported feature in their organization, compared to 50% a year earlier. A particular increase is visible in developing countries, where AI adoption increased by 12 percentage points from 45% in 2020 to 57% in 2021.
As the report shows, as many as 74% of respondents indicated an increase in sales thanks to the use of AI-based solutions in sales-related functions. 27% of respondents indicated that the use of AI is responsible for at least 5% of their EBIT.
The report shows general trends in the use of AI in all functions of the organization, from Production, to HR, Product Development, Logistics, Finance and Operations. Because of the area of AI that we deal with, I will focus on the area of Marketing and Sales.
Below, I list what are, in my opinion, the 4 most interesting elements of the report along with my key theses regarding them. As all the elements indicated are very extensive, I have taken the liberty of describing them in more detail in the following 4 articles. Links to these will be shared in future posts. Of course, I was looking at the whole thing from the perspective closest to my specific area of activity, in particular the application of AI-powered Image Recognition in supporting processes in Retail.
Why do I find these 4 elements of the report so interesting?
Well, for two main reasons. The first is the focus on the area of Marketing and Sales. In many cases, the information contained in the report strongly corresponds to our experience with implementations of AI projects or discussions with our clients. Of these, I have selected those that have the greatest impact on success and effectiveness in practice - and that is the second reason.
Additionally, it is worth noting that the report lists the two most popular AI applications in each area. In the area of Sales and Marketing, these are customer-service analytics and customer segmentation. In the first case, they can be any customer service tasks, including, for example, chat/video/audio bots and all related analyses. In the second case, activities related to collecting data about customers, including data supporting CRM or intelligent online shopping suggestions that are based on observed customer behavior (Amazon was one of the pioneers in this area).
These methods of AI application did not actually change that much compared to the previously published Capgemini report. You can find more information about this report here.
Importantly, in the area of Marketing and Sales, Image Recognition does not occur, which is quite symptomatic. AI continues to develop fastest in the area of online sales. This is where AI models are most pleasant to build because they can be based on an adequate amount of readily available data. In physical stores, where the quality of the display and the physical accessibility of products play a very important role, it is much more challenging to obtain such data, especially for the purposes of Image Recognition. The availability and complexity of the construction of such models are much more difficult. The topic of AI supporting Image Recognition in retail itself is a different story. Despite these differences in data availability and quality, the following 4 elements taken from the report are also highly symptomatic of Image Recognition.
I strongly encourage you to read the publicly available information that McKinsey has published on its website.
In the following days, we will be publishing next parts of the article:
Part I. Construction itself is not enough.Part II. The double increase in cost identification.
Part III. Experienced users are better at estimating AI costs.
Part IV. It is worth engaging users.
👉 Follow us on social media to stay up to date:
