What Role Will AI Play in Digital Marketing by the End of This Decade?
The growth of artificial intelligence has been significant, but not always understood. Yet it makes sense to assess how it might influence marketing.
The idea of artificial intelligence taking on important decisions has been a classic ‘marmite’ development.
For some, the idea has been a welcoming one, based on the notion that machine learning using levels of data that no human mind could absorb. It will be able to assess matters and establish facts at a hitherto unimaginable level. The result of this, it is argued, could be the production of data that will lead to more accurate and informed decision-making.
The alternative view has been that AI could ultimately have a negative human impact:
- The robots would know too much about us, which might be intrusive.
- Machines might even turn hostile (a concept much explored in literature and sci-fi, such as in the Terminator films).
- Machines could take away human jobs to leave millions out of work.
The latter idea has been taken seriously by key technology industry figures such as Bill Gates, who has called for robots to be taxed to help provide incomes for all those whose jobs would be lost.
Debates about the impact of automation on the wider economy will extend way beyond the sphere of marketing, but clearly it makes no economic sense to use technology to impoverish people. Quite apart from direct job creation – if there were no computers nobody would work in IT – the general trend has been for machines to help humans be in a more productive way.
What Has AI Already Done in Marketing?
When considering how the role of AI in marketing might develop over the coming years, it is important to start by appreciating that such technology has already been deployed in the sector. Examples of this include:
- Voice-activated customer service, for example Amazon’s Alexa
- Airbnb helps pair users with accommodation
- The use by Amazon and Tesco of voice commands to add items to a virtual shopping basket
- Personalised health and fitness advice delivered by a smartphone, used by Armour
Cases like these have become so familiar to us that many do not even see them as AI, but that is what they are.
Take the example of voice-activated devises such as Alexa: The capacity of a machine to learn to recognise human sounds has provided a facility that belonged to the realm of science fiction not so long ago.
Such voice-activated speakers will become increasingly ubiquitous over the early years of this decade, so by 2030 they may be as common a feature of households as smartphones and microwaves are today.
This will have its own impact on digital marketing, as all kinds of tools used – be it content, social media or PPC – will include some kind of voice-activated protocol.
For example, it could be content that includes a clickable link to a voice-activated device that could place an order, sign up for an email mailing list or request a call back.
What Could Big Data Have?
The use of increased data in the case of the Airbnb example is one way that AI can use machine learning to help fine-tune what people want. By recording and analysing the past experiences and actions of individuals, it can help to target them with the products and services most likely to suit them.
Already some supermarket apps do this on the basis of past purchases; for example someone who buys herbal teas a lot may be targeted with marketing messages when a new product comes into stock.
This may have several implications for buyer personas:
- At the outset, the use of AI to record consumer habits could help produce more accurate buyer personas. Through picking up emergent trends in customer habits faster might be the case through standard market research techniques.
- Buyer personas may be fine-tuned almost down to the individual level, with marketing being accordingly reshaped for specific groups in specific circumstances. An example of this might be people matching a certain persona who are experiencing a particular life event, such as moving home, changing job or becoming a parent.
- The persona may even be fine-tuned to take account of seasonal factors, such as how much someone might habitually spend on Christmas or summer holidays.
Could There Be More Limitations on Data Usage?
The use of big data and machine learning might bring marketers closer to understanding and being able to respond to what consumers want, but this also does raise some ethical issues about just how much marketers should be allowed to know.
Indeed, the issue of how data is handled has been the subject of various legislation, either domestic such as the Data Protection Act, or the GDPR rules introduced through the EU. Despite Brexit, there is no sign of any move to scrap the application of GDPR in the UK, so it must be assumed it is here to stay.
This poses a considerable unknown. How much can AI be applied in marketing using data before legislators decide it is too intrusive and act accordingly? While the growing adoption of technology – and perhaps the use of emerging new technology – could increase AI application, the use of legislation to constrain its use could act as a counterweight.
As a consequence, the mere possibility of further restriction on data use could cause marketers to be cautious in how they use data, lest these prompt lawmakers clamp down harder on what they are allowed to know.
What Else May Change?
While the development of technology suggests AI could have a larger role to play in marketing over the course of the next decade, it is hard to always be sure what exactly it will be for all manner of reasons:
- Recent economic and political history has shown that sudden and unexpected events can bring major change, from the 2008-09 economic crisis to the current COVID-19 pandemic.
- COVID-19 has made more people reliant on the internet, which could accelerate moves towards more automated and IT-based social and economic interaction.
- Disruptive technology may emerge that acts as a paradigm shift in how people act, much in the way that radio, television and the internet did.
For all these reasons, predictions should be treated with caution. However, what does seem clear is that AI already has some involvement with marketing, there is at least the potential for it to take on a vastly expanded role between now and 2030.