Strip away brand rivalries and there’s not much debate; AI-powered chat bots – what are becoming known as “chat engines” – have significant potential to upend the way we explore the web. That’s because they deliver information in a more iterative, direct, and human-centered way than tools, like search and answer engines.
They do more than threaten the business of search, though; they actually undermine the creator economy that fuels the data they’re built on. While “removing the middle man” may create a better user experience for searchers in the short term, in this case, it also disincentivizes those building content, drives data cannibalism, and promotes content degeneration over the long haul.
Whatever your feelings about how new AI powered tools work or whether they’ll replace search, they do represent a single inextricable truth. In an increasingly time-constrained world, tools which reduce friction in our ability to find, make sense of, and use the data that drives our lives are the only logical outcome.
For website owners who have relied on organic search to get found for the better part of two decades, its a scary proposition. It also means that succeeding in this new ecosystem will require a new skill: Chat engine optimization – or the ability to get the next generation of generative AI-driven chat tools to find and refer your organization as a subject matter expert.
The arrival of tools like ChatGPT surprised many of us. Even for marketing technologists, who are, all things considered, familiar with their evolution, a working prototype was years away… until it wasn’t.
While early versions of the technology have presented challenges, like the tendency to make up answers (“hallucinate”), many of them are already being addressed. For businesses who’ve traditionally relied on search to get found, it presents a larger, and looming, existential crisis.
“As long as humans still need a way to find the people, products, and services that drive our lives, chat engines will need a way to rank them.”
Search engines order third-party listing based on hundreds of unique “ranking factors”. The exact makeup of their algorithms are guarded but, after 20 years, we have the general idea. To that end, virtually anyone with access to YouTube and a couple of plugins can beat ninety percent of their compeitors in search for almost no cost.
Some would say that this democratization has led to the proliferation of spam content that’s made it difficult to find the answers one is looking for quickly. Others, that it’s at the center of an “economy of ideas” which has created data powering some of our biggest technological advances, including – notably – the large language model that powers chat engines.
On the other hand, chat engines use crawlers to aggregate data from around the web and generative AI to repackage it in their own language. While they do create a unique 1:1 connection between searchers and the data they’re looking for, they also seem to ignore the laws of supply and demand – feeding our own data back to us, often without even a citation.
In all but the most academic of use cases, that disincentivizes the person or business creating the data to continue doing so. After all, if sharing your knowledge and expertise doesn’t provide any net benefit for the creator, what’s the point in spending the time to create it?
The problem, it turns out, is that disruptive technologies disrupt. They do more than provide better solutions,, though. They open up conversations about how we accomplish our goals that involve pain, learning, and a departure from the known. A number of intellectual property cases shedding light on how we’ll handle these challenges are making their way through the courts now . It’s unlikely that any of them will reverse adoption of this technology though and chat engine optimization will become another strategy in most marketers discovery tool chest.
Had you asked us (or any other marketing agency) how to get found through search fifteen years ago, our advice would have included the development of “inbound” content that answers your customers’ most frequent questions.
In the spirit of ripping this band-aid off quickly, answer engines will likely have a catastrophic impact on these types of “What is” and “How to” strategies. Thought leadership, current events, and case study content will likely fare better since they don’t represent common knowledge. But, they have always been less intent driven and perform better on discovery-based social media platforms, anyway.
While we may yet end up in a world where chat engines replace search engines, the greater likelihood, at least in the mid-term, is a collaborative search/assistant similar to the one featured in Bing’s blended discovery model.
As these technologies converge, chat engines will be built directly into a device’s OS. After that, they will become the OS itself. Finally, we’ll get IoT connected “smart” personal assistants that moves between devices and understands every aspect of our personal style, from the way we speak to the things we’re interested in.
In truth, this progression will be messier in life than on paper and is full of, yet undiscovered, opportunities and challenges. While chat engines may be good at helping us sort through the open web, they’re terrible (so far) at making use of it. That includes actions like feeding us, tailoring a suit, or providing anything beyond machine-generated call and response.
That’s because there’s a lot more to discovery than search. As long as humans still need a way to find the people, products, and services that drive our lives, chat engines will need a way to rank them. Understanding how that they do that will be key to success.
Around 2015, answer-based technologies like Alexa, Siri, and Google Home were starting to gain traction. Heated conversations about how to optimize websites for natural language processing were common and many postulated the arrival of a “brave new world” of online discovery.
In the long run, the technology hasn’t progressed much further than a mobile assistant. But, we did end up with lessons that may be applicable to how we approach chat engine optimizattion. One of the most important has been that as tectonic a shift as new search technologies often feel like, they tend to represent a change in the user experience rather than a broad algorithmic shift.
For example, when we asked ChatGPT for hints on how organizations of the future might optimize themselves to be referred by chat bots, it provided a laundry list of surprisingly traditional best practices: Fresh content, backlinks, keyword optimization, et al.
So, like answer engines, what first appeared to be a sweeping change in discovery may boil down to something much subtler: The difference between getting found and being referred. With that in mind, what should organizations focus on to ensure continued visibility on search while, while tweaking their website to be featured by chat engines?
In a world where content is scraped for solely for aggregation, it will still be an important tool in establishing subject matter expertise. We should assume that content which is used to answer a question at the top of the funnel, also contributes to the likelihood with which the chat engines presents the root domain’s owners as an expert, at the bottom.
Structured data isn’t new, but chat bots’ iterative question-and-answer style user interface means that it has renewed relevance. Those who can answer potential first round questions about their products and services with machine-readable endpoints are likely to have their content served first.
Low hanging “how to” content speaks to common subject matter that chat engines can aggregate from many similar sources. Other forms of long tail content, like case studies, still speak to your knowledge, expertise, approach, process, and experience, but do it in a way that is more proprietary and personal. Chat engine optimization will rely more heavily on third-party reputation.
Earned media, reputation management, testimonials, et al have always been important to high click-through and conversion rates. In an ecosystem that focuses on a fewer number of higher quality and more relevant results, expect this type of social proof to become more relevant at the top of the funnel, though.
The ability to refine based on feedback is unique to chat engines. To do it effectively, they need a depth and breadth of information about your products and services that allows the AI to have this back and forth. It’s not just enough to get found. The websites best able to answer common customer questions in both human and machine-readable formats will likely win out. FAQs offer a breadth and depth of information about your business without clogging up your landing page’s storyboard.
At the end of the day, in may be helpful to rethink the way we explore the world’s data as being discovery rather than search, answer, or chat-based. Each will play an important role in helping us find and leverage the data we need and they’ll be increasingly deployed in concert.
For those that have relied heavily on the search ecosystem, it’s not all doom and gloom though. While chat engines will decrease the overall volume of traffic to our websites over the next decade they will also act as qualifying agents, improving the relevance of traffic we do get and ensuring more frequent and higher value conversions.
Chat engines will also reshape how we engage with technology at large. They will both humanize and make it more seamless, allowing us to move more fluidly through the world by providing just in time access to the answers we need for every aspect of our personal and business lives.
Omnichannel marketing offers communications leaders a powerful strategy for creating relationship-driven brand storytelling . It can be enormously effective at growing a loyal following, but delivering on that promise is often more complex than it first appears....