AH Marketing

View Original

So Long SEO Traffic: The Future Is AI-Based Search

The future is AI-based search, and it’s coming sooner than you think. From the artificial intelligence (AI) assistants we’ve had for years (Siri, Bixby, Alexa, Cortana, and Google Assistant) to the new Bing Chatbot, Microsoft implementing AI into Office, and the increasingly smart ChatGPT - we are starting to see where AI search may be headed.

Let’s take a peek at the latest search results, starting from when you ask a question and ending with the type of answer you’ll receive from Google, Bing and OpenAI’s ChatGPT today.

Today on Google 

Today, when you post a query to Google Search, Google may respond with an answer known as a Featured Snippet gleaned from copy mentioned within pages related to your questions taken from its index of trusted websites.

With Featured Snippets, Google displays the information a user requests on the search results page so that users receive answers without having to click through to a website. Featured Snippets have already earned the nickname Zero-Click amongst the search engine marketing community and employees at Google. As the nickname indicates, when a user perceives that their question has been answered, they are less likely to click on the website link provided at the bottom of the snippet.

Let’s now turn to Microsoft Bing search, which now includes AI-based search results alongside its traditional search results.

Today on Bing

For the user, the search process is similar, but Bing now utilizes AI to enhance what used to be solely driven by SEO and search algorithms. According to Bing Chat, “The new Bing runs on a next-generation model from OpenAI that is more powerful than ChatGPT. Microsoft worked with this OpenAI model to develop a customized set of capabilities and techniques to combine cutting-edge AI technology and web search.”

Today on Bing, when you conduct a search, if the chatbot recognizes it as an AI-directed inquiry, additional information will appear on the right side of the page with sources cited at the very bottom. This solution is much like Google’s Featured Snippets, only including an AI response and taking a slightly longer amount of time to generate a result.

Another option provided by Bing is Bing Chat, which allows users to prompt its chatbot directly. The bot will automatically reply and cite its sources at the bottom if you’d like to click on them for more information. Of course, the bot can also elaborate further and go into more detail when prompted.

With an answer so convenient and fine-tuned in this way, however, you wonder how many users will actually find a reason to click through to any of the sources.

Now let’s examine OpenAI’s ChatGPT, the popular program behind Bing’s AI chatbot.

Today on ChatGPT

First, consider my conversation with ChatGPT:

April 10, 2023

[When prompted with: “What features should I look for in running shoes when planning to run 5 miles a day, and what are some good options?”

ChatGPT responded with: “When it comes to running shoes, here are some features you may want to look for when planning to run 5 miles a day:

  1. Comfort: Choose shoes that fit well, feel comfortable, and have enough room in the toe box to allow your toes to move freely, and the heel should be snug enough to prevent slipping.

  2. Cushioning: Running shoes should have adequate cushioning to absorb shock and protect your joints. The amount of cushioning you need may depend on the surface you'll be running on and your personal preferences.

  3. Support: Look for shoes that provide the appropriate level of support based on your pronation pattern. You can have a professional gait analysis done to determine your pronation pattern and choose shoes that provide the appropriate level of support.

  4. Durability: Running shoes should be durable enough to withstand daily wear and tear.

  5. Breathability: Choose shoes that are breathable to help keep your feet cool and dry during your runs.

As for some good options, here are a few popular running shoes that are recommended by runners and experts:

  1. Brooks Ghost 13

  2. Asics Gel-Kayano 27

  3. Nike Air Zoom Pegasus 38

  4. Saucony Ride 13

  5. Hoka One One Clifton 7

It's important to keep in mind that the best running shoes for you will ultimately depend on your specific needs and preferences, so it may take some trial and error to find the perfect pair.”

I then prompted it a second time with: “Please cite your sources.”

For it to respond with: “Certainly, here are some sources that support my previous response:

These sources provide information on what to look for when choosing running shoes and offer recommendations for specific running shoe models.”

Based on this conversation, we can see that at its core, ChatGPT is less inclined to cite sources unless specifically prompted to do so. This is likely to continue to be the case with newer forms of AI. Once you have the answer, why consult the source?

This is important because, going forward, as many more apps and programs start to integrate AI for their purposes, there is little incentive to send users to external websites where the source material lies for the risk that the user won’t migrate back. Consider late 2021, when over 200 news publishers sued Facebook and Google for the way the social and search giants were profiting from news content at the expense of the publications that produced it. AI models seem to be doing virtually the same thing, and this could foreshadow legal implications on a vastly larger scale.

It’s also important to note that even though ChatGPT did cite its sources during my initial test, two of the three source links provided were invalid or led to a completely unrelated page. This means that even when prompted, ChatGPT may not always provide users who do want to look further with accurate sources for their inquiry.

Today on Bard

Bard is Google’s new AI language model, Google’s response to OpenAI’s ChatGPT. Bard is yet another language model built to generate text and answers to users’ inquiries. Contrary to its previous statements, Google claims that “Bard is not Search.” As we await the future AI integration into Google Search, let’s take a look at what kind of responses we can glean from Bard. 

From this example, we see that Bard is generating a similar response to those we receive from both ChatGPT and Bing Chat. However, Bard would not cite any sources, even when prompted to do so. Bard’s explanation was that “Bard and ChatGPT are large language models, not knowledge models.” Remember, however, Bing Chat automatically cited its sources, and ChatGPT is occasionally able to provide sources for its responses while Bard, also a language model, was unable to do so.

With a shift in focus away from elevating source pages to elevating answers and delivering app functionality, companies and their search marketing professionals (who have for years invested in building informative content to drive search traffic to corporate websites) should be aware of the implications of AI-based search and how it might potentially reduce website traffic and new user acquisition. In fact, the evidence also suggests the practice of search engine optimization (SEO) itself may be in danger of becoming extinct.

How Often Do You Peruse the References Section of a Book?

Google’s stated mission is to “organize the world's information and make it universally accessible and useful.” It’s important to remember that Google’s purpose is to serve the searcher, not the source. While AI may still take SEO into account when generating results, we are likely to see a decline in referral traffic to websites because of a change in the nature of how AI responds as compared with search engines. AI and its ability to access information from the web and other sources when prompted allow AI to leapfrog the role of SEO and provide answers directly to users sourced in part from the hard work and investments made by companies, content creators and thought leaders over decades.

A recent study by BrightEdge has found that organic search currently delivers an average of 53.3% of all traffic to websites, far greater than any other channel. In fact, SEO drives 1000%+ more traffic than organic social media today. But as AI becomes more integrated, what will these numbers look like six months or a year from now? We already see AI being integrated into various apps and websites and integration within third-party software such as Microsoft Office. With AI conveniently supplying answers within applications in real-time, search engines themselves may eventually become obsolete. 

We cannot rewind the clock. With traffic decreasing from search-based sources, how can companies begin to divert their efforts to other effective marketing channels and techniques ahead of rapid declines in search traffic?

The Future of SEO And Web Traffic

As AI continues to learn from our content and passes answers on to the consumer with little emphasis on sources, consumers will by nature come away from those engagements with less brand awareness as they acquire what they believe to be the best and most convenient responses to their inquiry. As such, brands are likely to have less organic influence on consumers by way of searches. Instead online purchases may increasingly depend more on advertising spend or factors beyond the brand, similar to what is already happening on e-commerce sites like Amazon and eBay.

The next phase of search will likely deemphasize, and in many cases, not natively pass along the conglomerate of sources that helped inform AI responses. So, how can we drive organic traffic in a post-SEO marketing world? Instead of focusing your company’s limited resources on generating as much traffic as possible while SEO is still effective, consider this marketing model instead:

  1. Talk directly with your customers.

  2. Determine exactly what’s standing in their way.

  3. Stand up for your customers in the market with a simple yet powerful point of view (POV)

  4. Evangelize the POV consistently across your customers’ most influential channels.

  5. Consider how social media, newsletters, live events and community building can help extend engagement beyond public channels.

Done this way, you won’t be reliant on search engines (and their constantly shifting priorities) for your marketing. Instead, you’ll control distribution and engage directly with loyal audiences of advocates who can pass your messages along to the people that need to hear them most.

Need help? 

AH Marketing is a full-service marketing greenhouse for companies that need a bit of sprucing. We support growing B2B businesses with marketing strategy, planning and execution that results in increased demand and pipeline. We want you to be wildly successful. Please contact us for help.

SUBSCRIBE to our LinkedIn Newsletter for weekly updates.