Is AI worth the hype?
According to Gartner CMO spend survey, an allocation of 27% of expense budget to technology, which brings an overall of 3.24% revenue in 2017, indicates an excessive dependence on marketing automation.
An adoption of artificial intelligence (AI) in marketing is reflected through auto-generation of personalized content, product recommendations, user identification, email campaign automation etc., within different industries.
An expontential growth in AI spending to $52.2 billion by 2021 with an annual growth rate of 46.2% within 2016–2021 period on a global scale is forecasted by the International Data Corporation.
A full integration of AI into the marketing ecosystem is undeniable but a quesiton that can be imposed is whether the input of resources and investment into technolgy is justified?
What does AI bring to the table?
Processing of big data that is being transformed into a predictive model based on users’ digital activity has been leveraged by marketers not only to generate personalized content and recommendations, but also to predict which target segments are more likely to convert with more detailed information on demographics, purchase history or preferences.
A projection of 85% of customer interaction being substituted by technology, such chat bots, reflects AI’s cost efficiency and lack of dependence on live agents to process a big volume of requests and decrease waiting time period. The limitations of technology-driven customer support systems can be found in its inability to handle case-specific inquiries, special assistance requests and due to its complexity, additional staff training, and maintenance might be required.
Customer retention and acquisition, in which a unique identification number is being assigned for further brand interaction and tailor message delivery for individual customer, are also AI tools being utilized by marketers.
Can AI reach level of perfection?
Is AI’s capability to process and analyze big volume of data in a timely manner that is beyond human capacity enough for marketing implementations? Despite the performance of routine marketing operations that involve automation of campaigns and cross-channel personalization, human intervention is still needed. Due to technology’s lack of creativity and cultural references, a generation of innovative and case-sensitive decisions creates boundaries for a full AI integration.
A formulaic creation of monolithic entity via mathematical algorithm overlooks a core concept of human nature, in which an existence of outliers, unique consumer behavior and preferences can serve as a barrier for machine learning that uses one-size-fits-all approach.
A potential intrusiveness and data privacy concerns due to personal information being applied within the ad content might be deterrents in consumer’s buying decision process.
Are big data and AI all time solutions for marketers?
According to the survey data, 63% of US marketing technology decision-makers attributed personalization of digital campaigns, data unification and customer profile maintenance as their customer data management solutions. 46% of survey respondents identified AI as the most important technology for business and IT decision-makers as of November 2019.The prevalence of marketing automation and AI utilization among advertisers pose a question on whether data itself can serve as a basis for decision making process for marketing implications.
Data value depreciation and irrelevancy given the annual data decay rate of 54% within the average customer databases would not be sufficient for making future marketing decisions. Data inaccuracy that affects report analysis due to 45% of web traffic deprived from bots is also a potential big data limitation. A “dark social” concept suggested by Mr. Madrigal in his research about The Atlantic’s web traffic analysis that revealed 56.5% of traffic coming from undetected sources had been categorized as a direct traffic, illustrates a significant occurence of data discrepancies.
Moreover, generation of predictions and recommendations based on statistical or heuristic analysis provides a descriptive analysis rather than explanatory examination that cannot be translated into actionable insights undermines AI and big data efficiency.
How would GDPR and CCPA reshape marketing landscape in the cookieless world?
As data privacy policies and regulations attempt to indirectly reinforce limitations on AI and machine learning, over 40% of US marketers and agencies determined sales-lift and ad effectiveness research as the most important ad measurements to be integrated as of September 2020.
Increased spending and emphasis on use of first-party data among 60.4% survey respondents were declared as a result of GDPR and CCPA regulations.
A gradual shift of ad tech industries in favor of fingerprinting, in which a digital trail is attached to the battery status or browser window size to track and assign a unique identifier code, might occur due to its proven effectiveness based on Princeton online tracking research on top 1 million websites.
Shall all the bets be put on AI or humans are still relevant?
Boston Consulting Group (BCG) study that evaluated a combination of AI, machine learning and human intervention in relation to the marketing campaign performance revealed that despite advanced technologies increasing performance by 20%, a human input was still required for strategic considerations based on three dimensions, such as alignment of campaign objectives and technology settings, adjustments of channels for category compatibility, and audience and message selection within customer journey stage. A human intervention and appropriate adjustments provided an additional 15% increase in the campaign performance, which implies that the combination of both would lead to the most effective and applicable results.
AI and machine learning’s dominance at iterative learning and human’s exclusive capability in target segmentation and more accurate perception of changing consumer behavior are not mutually inclusive and both would serve as assets to the marketing ecosystem.
