Boston Consulting Group (BCG), Gartner and McKinsey all predict that customer experience will be one of the fields to benefit most from advancements in generative artificial intelligence (generative AI). As part of a survey of business leaders, Gartner determined that the largest share of the investment in generative AI, some 38%, will be channelled into customer experience and retention (Gartner, 2023).
Just to remind you, generative AI is a technology that facilitates the production of various types of content, such as text, images, audio, video and data.
There are three aspects of customer-centric marketing that can be improved through generative AI: customer understanding, personalization of customer experience and the protection of personal data.
How can generative AI improve customer understanding?
In terms of customer understanding, generative AI facilitates the analysis of unstructured data (text, images, video, audio) in order to extract trends or insights. Unstructured data are generally hard to analyze, but with advancements in deep learning (or machine learning), new opportunities have been made available to marketers. Access to these tools has particularly been democratized in the past two years.
Here are a few applications of generative AI in the area of insight:
-
extraction of key themes from product reviews or comments/feedback in order to evaluate a service;
-
automated summaries of all market searches (secondary search/desktop research) performed in order to understand key segments, irrespective of the document format used to support the search (PDF, webpage, video);
-
analysis of survey responses to extract the main insights;
-
analysis of images or videos viewed by customer segments in order to detect trends.
These are examples of just a few use cases to illustrate the potential of this technology, but there are many other possible uses.
How can generative AI improve customer experience?
Generative AI will enable hyper-personalization of content throughout the customer journey using multimodal systems (the production of a variety of content in various formats: text, image, video, etc.). Web users will be increasingly exposed to AI-generated ads, from acquisition through to retention; the same will be true for content displayed (email, webpage, mobile screens…) on digital properties. The content generated could be based on the customer’s marketing profile (360 customer data or profile) in order to offer relevant experiences that are unique to each customer. This technological innovation will reinvent the relationship between consumers and brands (relationship marketing).
Content or product recommendation systems will be enhanced using generative AI so they offer choices (in an interactive mode) using language that is much closer to natural conversation. The ability to generate several variations of content around the same theme will facilitate and accelerate testing (A/B/C testing).
Customer support through virtual assistants will be updated through the possibility of engaging in richer conversations with consumers. This will also facilitate the generation of reports of these conversations (with recommendations) for the humans tasked with handling them.
These virtual agents could also leverage cross-selling or upselling if they are supported with reliable customer data. In terms of operational efficiency, a virtual assistant would be able to support customers 24/7. Although this technology is faced with certain challenges at the moment, its potential remains very real.
One of the big advantages of these innovations is the possibility of combining traditional AI models (purchase prediction, next best action) with generative AI systems to deliver even more advanced personalization.
The importance of capturing high-quality customer data is critical for achieving this level of hyper-personalization. The precision and reliability of these tools will be critical for successfully deploying such an ambitious initiative. Obtaining the consumer’s explicit consent and protecting their personal data are two challenges that must be taken into consideration in order to operationalize this use case.
How can generative AI improve the protection of your customers’ personal data?
Generative AI doesn’t just apply to text, images and video; these models are also used to generate data. One specific example would be a company that decides to generate synthetic data based on existing data in its customer database.
Synthetic data are artificially generated by a machine learning model. These data must maintain the same statistical properties as the original data to be useful. For example, if the original data set is composed of 60% women, the synthetic data set must also respect this ratio.
Gartner foresees that by 2030, most of the data used for machine learning and AI will be synthetic data. Sellers currently offering solutions for transforming real data into synthetic data are Mostly.ai, Datomize, Synthesized, Hazy and others.
The two biggest advantages of synthetic data are:
-
the ability to generate data sets randomly upon demand without having to collect data;
-
on-demand use of data without jeopardizing the privacy of your customers’ personal information. Synthetic data are de-identified; an individual could only be identified within a dataset with great difficulty.
From the perspective of customer-centric marketing, synthetic data can be used to analyze customer behaviour without the use of their personal information. Synthetic data facilitate the training of predictive or segmentation models that aim to improve consumer engagement with the brand.
This is an important point, because synthetic data are de-identified. They therefore cannot be used for precise ad targeting, because they do not contain any persistent identifiers.
I hope this article will inspire you with potential ideas for operationalizing customer-centric marketing using generative AI. If you need any further information, please get in touch with us!
