Search, small data, AI and the shape of things to come

My personal truth is that SEO and search-engine results pages (SERPs) don’t influence me much. When I set out for knowledge, the info on a SERP may lead me to click here and there, but the SERP process is slow and I am more likely to actively seek the information I need elsewhere.

SEO is, in my estimation, becoming a low-value standard, a must-have which is necessary but not sufficient. I will only use a SERP to pluck out specifics on a site when I am clear on what I need. If not professional knowledge, I am more likely to head to social media search to find a product that I need.

Use it or lose it... make sure that your social media is on point. Staff and event announcements may bring fleeting awareness and superficial brand engagement but there must be an overt or at least implicit call-to-action (CTA) that has a goal directed result, either in the form of sales, behavioral change or the psychosocial elements of transformation or tribalism.

How am I influenced? The answer lies in small data which is individual to me. It considers my online behavior, the tenor and content of sites I visit, as well as my online purchasing behavior. For instance, I buy e-books that Amazon recommends to me on its site and in emails books based on my previous purchases. Google Cards (swipe left to right on an Android device) knows the type of movies I enjoy that I have interests in cosmology and small-form NUC computing,

The benefit of having relevant content delivered by algorithm, specific to me, outweighs the need for personal privacy. I knowingly train algorithms and use the opportunities that they give me to refine and correct my algorithms.

AI language models (LMs) are currently large language models (LLMs) which use vast data sets to aggregate responses, however there is a shift to the idea of smaller LMs customized to brands or individuals.

The brand / enterprise-specific AI is self-explanatory. Based on prompts these could deliver either brand and product information or in an exclusive instance of AI, internal enterprise information (similar to an intranet) but with the added benefit of productivity support.

An individual AI is more intriguing, the grail of current development. GPT reports (in a response to a query that I gave it), that there is progression towards this goal. The uses would be personal productivity in the workspace, maybe mirroring the internal enterprise AI but, most importantly, pursuits such as personal interests, personal creativity, unique consumption behavior, brands which will earn interest and loyalty and a speculative component of what may or may not become an interest.

The key is ‘persistence’, the ability of the AI to recognize you and come back with responses that are useful and of interest to you but that also recognize that you evolve and cater to that. For instance, most AI image generators do not recognize and repeat a figure in an image. Characters vary from image to image. The idea of persistence corrects this and will do so on an individual basis for other types of pesonalized information and content.

Persistence is not device specific nor site specific like Android cards or Siri. It will recognize your changing environment, character, personality and biases. Its algorithmic nature will be determined and managed by you.

This is the future of communication and content. It is still unclear how it will roll out and how brands will feed in, but it is the future that faces brand management, one that needs to be prepared for.