Nuance is a funny thing. Intangible. It comes so naturally to most of us, we don’t notice it until it’s not there.
Each human experience is shaped by our socialization. Sure, we have in-built traits and temperaments at our core, but the belief systems we ascribe to and the biases that make us fallible shape what we become. Nature vs nurture.
We’re at an analogous inflection point with the use of AI in Search Engine Marketing (SEM).
AI is far from ‘new’ in search. IBM’s Watson Analytics for Content, released in 2015, paved the way for AI SEM tools. But the huge acceleration in, and acceptance of, AI tools in the last two years has meant marketers increasingly understand its nature, but are still some steps away from nurturing it to best use.
Marketing has long been a balance between art and science. But in SEM specifically, as AI moves from simply an insight tool to a channel in itself, it’s a critical time to get to grips with what the technology can, and can’t, do for search strategy.
Search is set to change
AI, unsurprisingly, was a hot topic at last year’s Google Marketing Live, with the company sharing new details around its Search Generative Experience (SGE).
SGE is set to transform the search experience through generative AI (AI that learns patterns and makes predictions). It will allow users to get more tailored and insightful information from search queries by asking more complex questions, getting to the gist of a topic faster, generating drafts and imagery from searches and making faster progress through a topic, by asking follow up questions to fluidly refine responses.
It’s an important evolution in the relationship between artificial intelligence and marketing. So, as AI in SEM enters its formative years, here’s a look at what comes naturally to the technology, where capabilities still need to be nurtured, and some ‘watch outs’ for ensuring nuanced collaboration between marketer and machine.
Strategy & Planning
Nature: Smart Bidding has been a mainstay of the marketing mix for a number of years, and Google Broad Match has evolved quickly to better analyze customer intent and predict outcomes. AI powers crucial insight tools, such as Keyword Planner, Looker, Tableau and HotJar, to name just a few. By leaning on AI-generated discovery, humans are freed up to spend more time understanding data and making evidence-based decisions.
Nurture: AI cannot integrate and rationalize the specific organizational goals, real time market conditions and subjective opinions that must all be considered in a robust strategy. It doesn’t have the real world experience or audience intimacy to build subtle distinctions into brand and campaign planning. Marketing leaders also remain unconvinced by AI-generated SEM predictions. The tool needs to be trained and nurtured on larger and more varied datasets to better grasp points of difference and project outcomes.
Nuance: There’s still a strong need for humans to lead and oversee strategy and planning. People will be required, at least for the foreseeable future, to bridge the gap between the data that AI can produce, and the actual insight and action to be taken from it, according to the conditions of their own organization.
Creative
Nature: AI can create thought starters for humans to consider, build on and refine, and it can cut research time drastically by streamlining information from multiple sources. Dynamic ad serving makes it a huge enabler for personalization and we’ve seen some gorgeous, true to life imagery generated, given the right prompts.
Nurture: AI cannot think or feel like a human, so its ability to change attitudes or shift behaviors is limited. It can’t understand feelings or empathize with lived experiences, so its ability to develop creative or copy that makes emotional connections is stunted. Importantly, AI-generated creative can be biased and potentially socially divisive, so organizations need to tread very carefully when letting AI loose on brand visuals.
Nuance: To overcome bias, AI needs to be trained on diverse datasets. According to Google, ‘The data that SGE is trained on is based on high-quality web extracted data that can exhibit narrow representations of people or potentially negative contextual associations.’ AI can also let subjective opinion leak through, and needs to be trained to better understand where a personal perspective is coming through in the source materials, and when it is or isn’t appropriate to cite it.
Delivery & Optimization
Nature: In terms of delivery and optimization, AI is already having, and will continue to have, a huge impact on SEM. It analyzes campaign performance and adjusts settings in real-time and without rest, uses NLP to understand intent and determine which search terms are most likely to lead to conversion, and can run A/B tests and implement learnings, once parameters have been set.
Nurture: AI is only as effective as the information and guardrails on which it operates. For example, an AI can optimize for ROI, but the onus is still on marketers to determine and program the value and importance of different conversion events.
Nuance: While we can confidently let AI take the reins in campaign optimization and delivery, human involvement remains vital for strategic decision-making and real-time data analysis. A good example of this is in high-pressure or high-demand scenarios such as Giving Tuesday. Conversion lag within platforms means that ROI can appear deflated in real-time, necessitating that marketers bridge the gap through analytics or CRM platforms.
Measurement & Reporting
Nature: AI has a huge in-built advantage over humans in measurement and reporting. It can rapidly analyze huge datasets, visualize findings and empower people to make faster, genuinely evidence-based decisions. A human brain doesn’t stand a chance against this capability.
Nurture: Despite AI’s vast capabilities with data, there’s still a strong reliance on strategists to glean the insight from the numbers. AI may overlook nuanced insights and patterns that only humans can identify through experience, expertise and strategic thinking.
Nuance: Measurement and reporting has long relied on AI as its foundation. However, it is still up to the human to turn this information into something that makes strategic sense and lands with stakeholders. Marketers will become increasingly liberated from data science, but the time returned will need to be spent interrogating, translating and making sound judgments in the strategic interest.
Evolve your approach
It remains to be seen whether AI will ever be able to replicate the subtlety and subjectivity intrinsic to the human condition and therefore, to effective, emotionally-intelligent marketing. Nurturing it towards these nuances without building in our own biases, and into a technology mature enough to replace human marketers is a long way off, but SEM gives us a particularly good insight into what AI can do—in collaboration with humans—across the marketing tech stack.
