How to Use Artificial Intelligence in Marketing

How to Use Artificial Intelligence in Marketing

Artificial Intelligence (AI) is a disruptive technology that will impact a diverse range of industries. Even though the majority of the marketing managers are conscious of the ground-breaking potential of AI, many of them are unaware of its benefits or of how they can adopt AI to upgrade their marketing. To put things in perspective, only few marketers actually use the advanced capabilities which AI offers like personalizing campaigns, collaborative filtering and predictive models.

However in today's times, because of the availability of data and modern computing power, AI can garner insights that go beyond those of traditional statistical methods. Multiple customer-focused applications across a variety of fields are now possible.

ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING
Classically, AI is a technology that enables a machine to learn and perform human-like functions. Machine Learning (ML) software can detect patterns and make predictions and recommend solutions by processing the data rather than by inserting explicit programming instructions.

Applications of AI in Marketing
Here is an organizing framework to help modern marketers understand the practical uses of AI for strategic marketing:

CURRENT SITUATION ANALYSIS: This step involves a fundamental understanding of the macroeconomic factors that directly affect the organisation, the stakeholders and the marketing. AI methodology, including social listening, can draw conclusions on markets and consumers, including customer satisfaction, purchasing patterns and demand analysis. Current applications include; understanding unstructured data, anomaly identification, sentiment analysis etc.
UNDERSTANDING CUSTOMERS AND MARKETS: Here, the marketers aim to understand the specific market they wish to operate in and the customers they want to target. AI can mine customer preferences from the web data, social media and mobile activity, which could then be analysed to provide feedback and take immediate actions. Current applications include; Market research using facial expression, eye movements, and audio comments, etc., mapping customer journey, including touchpoints (ads, influencers) etc. Medallia is one such consumer-experience software.
STP ANALYSIS: This involves developing and understanding customer Segments and then Targeting them with effective Positioning. AI assists marketers to predict customer intent and further segment them into more refined groups. For example, Harley-Davidson used Albert (developed by Adgorithms) to help automate and simplify marketing planning, which was done by providing past customer data. The data enabled the software to create similar audiences and match potential customers who resemble the current buyer. This software was responsible for 40% of Harley-Davidson's motorbike sales and also led to a nine-fold increase in inbound calls.
PLANNING DIRECTION, MARKETING SUPPORT AND OBJECTIVES: This stage involves making long-term goals and develop short-term objectives to support more prominent strategies. AI and chatbots can be integrated into apps or social media for encouraging customer purchase. For example, the Starbucks Barista, which is available on the Facebook messenger, enables you to order coffee via voice command or messages. The use of AI in customer service is projected to grow by 143% from 2019 to 2023.
PRODUCT STRATEGY DEVELOPMENT: At this stage, marketers have to make decisions involving the product design, quality, features and customisation. The opportunities for AI include identifying gaps for the development of a new product, production of customised products and also assisting product delivery and logistics.
For example, Lily AI encourages the customer to 'complete the look' at the checkout, i.e. it suggests a real-time head-to-toe look suggestion enabling the retailers to increase the basket size at checkout. Levi's uses AI to optimise product availability at their stores and improve size availability. Nike uses clustering algorithms to advice which products should be displayed together at the store. And Samsung uses Crimson Hexagon's AI-powered insights platform to understand how customers interact with the products and thus making marketing campaigns that the customers can relate to.

There is a range of opportunities that AI offers in marketing, but there is still a lingering fear that AI will replace human work roles across different professions. But to look at the broader perspective, AI will offer automation that will enable marketers to invest more in creativity rather than the process. Thus, regardless of the speed of AI adoption, future AI-enable marketing techniques will have and require both AI and human intervention.

⇒ What are your ideas and experiences with AI in the field of Marketing?

Source: Campbell et al., (2019) "From data to action: How marketers can leverage AI", Elsevier Inc., 2019, March-April 2020, Vol 63, pp. 227-243

Comments by date▼

Comments by rating

Rating

Chloe Xu
Director, Australia

AI Platforms, a New Battlefield for Marketers

AI (Artificial Intelligence) is a set of technologies, ranging from machine learning to natural language processing, which allows machines to sense, understand, act, and learn. AI is believed to transform the relationship between people and technology, and the way people live. AI platforms that host an AI Assistant, such as Amazon Alexa, Google AI Platform, and Apple HomePod, are being adopted by millions of households and play an advisory role of assisting their users to make various life choices.

Over the next decade, with technology firms and others fight to develop the preferred customer AI platform, AI Assistants will profoundly change how companies and customers connect. AI platforms will become the primary channel people gather information and get products and services, and marketing will turn into a battle of their attention. These changes will have a profound impact on brands, retailers, and marketplace dynamics.

Impacts on Marketing Practice

The changes mentioned above will impact marketing practice at three levels: customer acquisition, satisfaction, and retention.

Customer Acquisition
With customer data being analysed and used to create finely targeted marketing campaigns, customer acquisition will become even efficient. Customer acquisition will be more of a science and focus on a sole channel, the AI platforms, instead of on the existing multiple channels. Given the intermediary role of AI Assistants, marketers will need to influence the platforms' algorithms rather than the consumers directly. Brands and retailers should understand how AI Assistants prepare and present product information and make small purchase decisions for users on some occasions. Besides, like the current practice, businesses will have the opportunities to bid on or pay extra to be listed on the information presented by platforms. This feature will become an important revenue source for platform runners, and marketers need to shift their budget from advertising, listing, slotting fees, and retail commissions to AI platforms.

Customer Satisfaction
Having AI Assistants in the middle of companies and consumers, reliable customer satisfaction data would come from AI platforms than from consumers themselves. The platforms will predict what combination of product features, performance, and pricing is most appealing to an individual customer at a point of time. They may better understand and be enabled to satisfy customers than the customers themselves. Using AI platforms will lead to more efficient product sorting and matching in the market. Therefore, brands need to reinforce their positioning in ways that AI platforms can understand.

Customer Retention
AI Assistants can tirelessly review and compare all the offerings available on the market, which makes brand switching a lot easier for customers. This routine of revaluation offers opportunities for newcomers to the market since brand recognition plays a less significant role for AI platforms. It also forces incumbents to continuously innovate or adjust their product offerings to remain the market share.

Implications for AI Platforms

AI platforms can only succeed if customers trust them. Therefore, platforms need to ensure three things: accuracy, alignment, and privacy .

Accuracy. AI platforms function by offering a solution to their users. They ought to predict the customers' needs and give out a solution that can serve customers better than what chosen by the customers themselves. Otherwise, it is hard for the platforms to build a trustful relationship with their users.

Alignment. How AI platforms handle with paid and unpaid recommendations can either promote or undermine customer trust. Once users detect that their AI Assistant is pushing paying brands that do not align with their needs, they will lose trust on the platform. Keep transparent is the only way to get this problem right. For example, platforms can let their users know which is a paid recommendation and which is not.

Privacy. AI platforms use customer data to create tailored solutions. The more personal information they collected, the more accurate the platform. However, platform users may feel exposed and do not have any privacy. One strategy to solve this problem is to offer customised privacy setting, allowing users to control what information to share and how. Another less effective strategy might be arguing that privacy is protected as the data is processed by machines without human intervention.

AI platforms seem to be the future of sales and marketing, an area that the effect of brand recognition faded, and traditional retailing loses. But the good news is, 1) Brands can maintain direct connections with consumers as smart products now provide a channel for communicating and collecting data from the customers, and 2) 90% of global sales occur offline and for the foreseeable future, this situation will continue. So marketers still have time to understand AI platforms and search for ways to influence them.

Source: Dawar, N. and Bendle, N., 2018. Marketing in The Age of Alexa. HBR, 96(3), pp.80-86.