Attitudes to AI, a technology that has actually been around for decades in various forms, veer from apocalyptic to the wildly optimistic.
It’s a fast-moving world and in the past 12 months we have transitioned from a situation where analytical AI helped us find things that already exist, to generative AI where we can create new things that don’t exist.
This has potentially huge implications for marketing in areas such as image generation, text to image applications, or text to voice.
Will this put marketers and agencies out of work? I think not. You would never ask an accountant not to use a calculator to do your books, so why would you want to hamper your marketing department by depriving it of one of the most powerful tools they will have at their disposal.
AI is best thought of as a tool that helps them do better and faster work, but it won’t do it for them. I think of it as a calculator for the mind. The future will be about combining man and machine in a way that gets the best out of both.
Here are five ways GPT could do just that.
1. Just-in-time marketing
Whatever the tech involved, marketing is still about the right message to the right person at the right time and AI promises a level of hyper-personalisation because it will allow us to marry huge datasets with the meaning we can extract using GPT models; combining information with knowledge.
But one of the biggest potential challenges of this AI-powered hyper-personalisation is an old one – ensuring data privacy and security. Brands should be well aware of their responsibilities in these areas and resist the risk of over-personalisation as they explore these new opportunities. Don’t get ahead of what customers feel comfortable with.
The good habits brands have developed in delivering digital experiences that comply with privacy regulations are a good steer to keep AI use ethical and respectful. Be transparent about data collection and usage; obtain explicit consent; provide customers with access to their data, and let them opt-out.
2. Insight automation
Marketing is all about gleaning the golden insight from the mountain of data and AI algorithms can make this happen at warp speed, cleaning and organising data, analysing and unearthing patterns and trends. Crucially, it frees up time to focus on more strategic tasks, such as identifying new opportunities and developing creative strategies.
A GPT model can analyse client lead-generation data to see which channels are generating the most leads, and which segments of your target market are the most responsive to each channel.
Inaccurate predictions and unreliable insights can result from poor data quality or algorithmic bias – a reminder that we can’t simply step aside and leave it all to technology.
It’s also imperative you don’t use confidential data on public tools where data is re-used for future learning cycles. As a rule of thumb, only use public data or information on public tools.
One caution is that it’s not just the data you submit as part of the question. Because ‘meaning’ can now be extracted, the phrasing of public information can imply confidential knowledge.
Fortunately, AI providers are aware of this issue and very much like it exists in other technology like cloud computing, there’s way to use AI services in a segregated way, where you opt-out of the services being able to use the data for future modelling.
Microsoft already has entreprise cognitive API available set up in this way and OpenAI recently announced it will launch business versions of its models.
3. Enhanced customer experience
Until now, a lot of the digital experiences so far have been based on an “if this… then that” type of interaction. For example, if a customer clicks on a particular link, they will see a certain display of information, as is often the case with chatbot “conversations”.
AI promises a more conversive and “intelligent” interaction rather than a robotic script thanks to the GPT models we now have. The new challenge for brands will be to control the narrative of these conversations. For instance, you don’t want the bot to suddenly promote your competitor.
With smart use, AI could foster stronger relationships through better customer experience providing more personalised support, faster response times, and more relevant content. Not only will this be available instantly and around the clock but predictive analytics will anticipate customer needs and provide recommendations in real-time.
4. Productivity gain
The biggest obvious benefit from AI is automating routine tasks such as creating email newsletters, social media posts management, generating reports, conducting keyword research, or managing a database. However, human creativity, empathy, and critical thinking are crucial so you need to strike a balance that leverages AI’s capabilities while keeping human oversight and decision-making in the marketing process.
When it comes to more strategic and creative work, AI can assist in the “diverge” stage of thinking, producing a wider range of possibilities to explore before “converging” on a key insight or idea. Again, it comes down to the capability to analyse large amounts of data quickly and accurately to uncover patterns and connections, leading to more informed and innovative ideas to create campaigns that really hit the spot.
5. Improved analytics and insights
By analysing vast amounts of data, AI-powered tools can identify trends, patterns, and correlations that would be difficult or even impossible for humans to detect. These insights can then be used to inform marketing strategies.
GPT can analyse customer data in various ways. It can perform sentiment analysis on customer feedback, cluster customers based on behaviour, provide personalised recommendations, and more. GPT-4, the latest version of GPT, can analyse images to identify patterns and trends in visuals, and allow marketers to discover which images resonate better with their target audience and optimise visual content accordingly.
Whatever the benefits AI brings to the practice of marketing, it seems unlikely that it will mean skilled professionals stand aside. Like all tools that have been developed over the years, some people will use them better than others and bad workmen will blame them for shoddy outcomes.
For those who master the tech, a fascinating future awaits.