The approaching shock wave is not from the noisy debate about AI authoring content. That’s a distraction from the seismic shift that has already been triggered.
AI is no longer just an author. It is also the audience.
How? Imagine you are a business seeking leads or suppliers in the quickest and most convenient way possible. Traditionally you may begin with a search on Google, then review the results. Recently businesses have employed a new strategy...
Businesses now instruct teams of AI agents to carry out search and ranking tasks for them, saving time and reaching a much greater scale.
AI agents can read thousands of pages of marketing material (written by AI or not – its inconsequential), compare it, summarise it, rank it and recommend further action, either for purchasing or sales lead qualification.
Sound far-fetched?
No, because our team has already deployed such teams of AI agents on behalf of multiple businesses. See our previous blog:
And its not just businesses. Consumers can also enjoy the convenience of agents searching the internet for them.
OpenAI launched SearchGPT, 1 Nov 2024
Perplexity.ai (Jeff Bezos) launched Shopping, 19 Nov 2024
Go to Market = Go to My Assistant
What this Means for Marketers
The AI agent has a different agenda to a human reader. It is not susceptible to clickbait or pretty pages, it is paid to carry out a specific task. Therefore:
Agents up rank the article which delivers the details sought
Agents down rank snappy articles that have edited out those details
Agents filter marketing content, removing extraneous content, videos and ads
Agents summarise content, the wording of the message is not preserved, only the gist
This means marketers must provide content for both the human audience AND for the AI agent. There are now two masters to serve. The table below sets out what that looks like in practice:
Audience: Human | Audience: AI Agent | |
Reader's objective | Scan quickly for relevance, engage emotionally, build trust through style and tone | Extract and analyze specific information, identify patterns and relationships, validate claims against criteria |
Where to publish | Social media, blogs, email newsletters | Company websites, no login required. Social media platforms often block bots. |
Content length | Concise, scannable chunks with headlines and visual breaks | Comprehensive and detailed. |
Supporting elements | Images, videos, infographics, pull quotes | Structured data, citations, verifiable claims, linked references |
Format priorities | Visual appeal, readability, mobile-friendly | Machine-readable, consistent structure, standardized formats |
Success metrics | Engagement time, shares, comments, emotional response | Information extraction rate, relevance scores, verification success |
Follow-up action | Click through, subscribe, contact, purchase | Data extraction, comparison against criteria, automated scoring |
SEO
Search engine ranking remains important because under the hood these agents use Bing and Google. Note that Bing becomes more important because AI agents currently pay to search, after all they don't read adverts. Bing is cheaper than Google.
Curious Return of Company Web Sites?
When we tested our agent team for clients we immediately noticed the team was more likely to find content when it was hosted on company web pages.
Social media platforms tend to ban automated readers and require logins which agents don't have. Some searches can get through these walls, but extensive content on a company page is more likely to give the agent what it needs than snappy social media postings.
Yes, that's the opposite of the trend in recent times.
Just Another AI Hype Cycle?
It’s early days and most marketers will not be targeting this audience yet. In consumer scenarios social media for human engagement will remain the priority. At least until we all have our own personal AI assistant.
But in the business context we want to cover large amounts of materials and get to the answers quickly. This is the point of an 'agent', it works efficiently on our behalf.
There are limitations. Agents can make mistakes when summarising articles, just like people. They need to be more reliable before being used to summarise medical data, but in many other fields the convenience of agents will outweigh the occasional error.
So in the short term we'll be writing for multiple commercial audiences. Many of us already do. We publish on the company page and put a short summary on LinkedIn.
In the long term, there may be scope for an industry which optimises content purely to capture the attention of the AI agents, just as courtiers would appeal to the adviser closest to the king.
Agentico is among the first Agentic AI advisors in the UK, supported by 10 yrs in Machine Learning and 20yrs in analytics.
Make sense of, and leverage, the seismic changes that AI agents are bringing to marketing or your industry. We're happy to talk more about the opportunities of Agentic AI and ML at your organization or event.
Get in Touch
Agentico.ai: AI with humans in mind |
Comments