In February 2024, Gartner published data suggesting that by 2026, Search Engine volume will decline by 25%.

I’ve held off mentioning this statistic in any articles or expressing a viewpoint on it, becuse I’ve been cautious about over-investing in predictions since ComScore published the “50% of all searches will be voice by 2020”.

Back in February, AI Overviews were still a beta Search Generative Experience (SGE), and a lot of AI tools and chatbots (such as ChatGPT) were being used by early innovators and early adopters.

But we now have AI Overviews globally, Meta AI pressing ahead with a launch in all territories, and the global rollout of Amazon’s Rufus. Google is also further the AI tooling in Shopping search with rollouts of things like Virtual Try On (VTO) and brand personalization, all touted in Google blogs earlier this year.

On top of this, Apple is pushing Search Intelligence as a unique selling position (USP) for its latest model, with mostly all Android OS phones pushing Google AI tooling, as well.

We’re closer than we ever have been to AI tools and chat bots crossing “the chasm”.

The AI Adoption Chasm

“The Chasm” is an idea from Geoffrey A. Moore’s book Crossing the Chasm. It explains the gap between two groups of early users in the technology adoption process: the early adopters and the early majority.

The mass market normalization of AI product features, communicated as being experience enhancers, normalizes and increases acceptance of AI with a larger audience.

This market readiness can be measured with the Overton Window, a concept taken from the political world. It helps us understand if proposed policies or political activities would be deemed as unthinkable, somewhat accepted, or universally accepted.

The Window is a continuum, as concepts can change over time. But we can apply this to emerging technologies such as AI.

A study by KPMG shows that attitudes toward AI are still “developing”. After surveying 17,000 people across 17 countries, they found:

  • Three in five (61 percent) are wary about trusting AI systems.
  • 67 percent report low to moderate acceptance of AI.

We can also look at other technological disruptors and how adoption rates have changed in the mass market as ideas shift towards the center of the spectrum and increase in adoption.

In my opinion, a good example of this with some similarities would be the adoption of mobile phones by the mass market.

The adoption and market acceptance (as shown by purchases) grew exponentially in the 20-year-period of 2020 to 2020 than it did in the period 1980 to 2020.

While the adoption of AI is happening at an accelerated rate, AI has faced similar uncertainties.

Mobile phones were met with cost barriers and some negative press around the potential health implications of using them. Similarly, AI has seen cost barriers (which have reduced significantly) and similar negative press around privacy and personal data implications.

Other modern technologies going through this same period of acceptance include autonomous vehicles and delivery robots.

The Move To Multi-Modal & Discovery Engines

Multi-modal search refers to the ability to query using multiple types of inputs, including text, voice, images, and video.

Search engines and social media platforms have been integrating these input types for a number of years, and AI seems to have accelerated this further.

A few examples currently available to the user include:

  • Google Lens (Video & Image Search)
  • Circle Search
  • Virtual Try-Ons (VTO)
  • Voice Search
  • Apple Intelligence
  • Microsoft Copilot
  • Meta AI

What makes other platforms so appealing to consumers is that they meet a certain number of adoption criteria.

Consumers adopt technology that they believe will improve their daily lives, increase efficiency, or provide significant value, whether in terms of productivity, entertainment, or convenience.

If the technology is intuitive and easy to learn, it lowers the barrier to entry. Complex or difficult-to-use tech can deter adoption, especially for less tech-savvy users.

Consumers are more likely to adopt technology that they perceive as safe and reliable, especially when it comes to handling personal data or online security.

TikTok

TikTok is proof that users who fall into the early and late majority in an adoption curve are willing to adopt “non-Google” technologies as answer engines and shopping portals.

For a while, TikTok was defined as a search engine. But it’s not; it’s a discovery engine. Social media is evolving from being just participation platforms to being engaging, multi-purpose portals full of user-generated content (UGC) curated by your immediate circles of influence.

SEO Data Practices Provide Insights

When it comes to using data in SEO practices, the primary goal is to harness insights that can inform better decision-making, optimize performance, and predict trends.

Our main function is to identify and create meaningful brand touchpoints at the point users are searching. The difference between SEO now and ten years ago is that users are searching for information on more than just Google.

This doesn’t mean that the “SEO data process” is now redundant. If anything, our processes are becoming more important and vital to the overall marketing ecosystem.

Users are still looking to go online to satisfy a certain need.

In my efforts to explain this to clients and in conference talks over the past year, I’ve outlined four user types who take to the Internet:

  • Researchers
  • Shoppers
  • Participators
  • Buyers

Researchers need to satisfy an informational need. We’d characterize these as top-of-funnel searchers or informational queries. These queries often vary in terms of openness and simplicity and can be formed on any platform.

At the opposite end of the spectrum, you have participators and buyers, who will go on the internet to make a purchase or to participate in some way, such as a peer-to-peer (P2P) forum. While AI can be active in this user journey, at present, they aren’t able to satisfy the user’s reason for going online.

Shoppers are somewhere in the middle. They’re likely researching and may be ready to buy, but not usually within a single instance of going online. They will swing between open and closed queries, ranging from simple to complex, until they finally make their purchase (or not).

SEO’s Evolving Function

SEO as a practice needs to evolve to maintain a seat at the table and ultimately drive value for businesses.

AI is a zero-data environment.

How can you define campaign key performance indicators (KPIs) other than “being present in AI Overviews” or influencing the AI generative response of any given AI platform when we can only theorise how and what the selection process is to appear in them?

Appearing in AI Overviews is as much SEO as it is SEO guesswork.

We need to communicate new KPIs that are better aligned with the business objectives and work in a way that works for the channel.

An example is SEO return on investment (ROI), which is always a contentious topic.

Paid search will always have a more traceable ROI than SEO. But when your organic performance stops working, it’s arguably a lot more costly to the business than if Paid is turned off.

Three rules of measuring Organic metrics in the AI era:

  1. Organic sessions may not directly lead to a conversion, but they can be valuable touchpoints along that journey.
  2. User conversion paths will vary between sectors, products, and the level of sign-off and consideration required on the purchaser side.
  3. Wider brand perception and existing customer sentiment (reviews) need to be acknowledged as being a significant influence on conversion likelihood.

This means we need to align our KPIs and success metrics closer to overall business metrics and more clearly show how we (both as a channel and a vendor) contribute to overall business success.

GEO & ODO

To optimize websites for the era of AI, the industry has coined the term GEO, Generative Engine Optimization.

But, given the prominence and rise of users’ will to discover and make purchases through the likes of TikTok and Instagram/Meta, we also need to consider ODO, Online Discovery Optimization.

Users are moving away from a single Internet entry-point ecosystem. Much like Netscape Navigator was the doorway in the 90s, Google and Facebook owned the “doorway” spaces for some time.

Now with multi-modal search and users accessing apps and other platforms directly, users are moving away from Google being the starting point in the journey.

This doesn’t mean other platforms are becoming “search engines”, but the way people seek out information online and how they form their journeys is yet to deviate from the established path. There are just more technologies and touchpoints in the mix.