How to Manage Online Reputation in AI Answers
It's not about what people say about your company - it's about what AI says people say about your company!
Online reputation management now not only includes the opinions of people, but also the opinions of AIs. Tracking your reputation in AI Answers will become as important as tracking your reputation in Search results.
Why does your online reputation matter?
"Reputation is a social construct based on the opinion other people hold about a person or thing."
Throughout evolution status has been a useful signal crucial for survival. Knowing who to hang out with made the difference between surviving the winter or not. Hence we are all descendents of people who paid attention to these reputation signals.
Since humans in general are status-seeking monkeys and online reputation is an easy way to assess statuses, then we are naturally drawn to the digital versions of these breadcrumbs as well.
During the times of print media, these status signals were compressed by news agencies and journalists who we assumed to be competent. Growing popularity of the internet however introduced voices of large groups of consumers who despite their assumed incompetence were able to collectively provide higher quality status signals than media with all its hidden incentives.
Whether you call it wisdom of the crowds or PageRank (the algorithm behind google counting incoming links to websites) doesn't really matter. What matters is that reputation is a biologically wired curiosity that's not going anywhere.
What does AI have to do with online reputation?
Reputation management is the practice of attempting to shape public perception of a person or organization by influencing information about that entity, primarily online.
If you look at the history of reputation tracking, then you'll likely notice patterns that hint at the logical next stage of technology in this domain. Let's look at how the spread of information evolved to understand why opinions and answers by AI matter and get a sense where things are headed.
-
In the olden days word of mouth allowed a person's reputation to spread relatively fast in close groups, but quite slowly across larger distances and larger groups. For what it's worth at the time the opinion of distant groups did not really matter.
-
With written word we started seeing compression and aggregation of opinion and reputation from multiple people into books and publications that had a much higher potential to spread across larger distances. As the world became better connected through trade the reputation of more distant groups became more relevant.
-
With the age of the internet the world was finally able to communicate at the speed of light. As we became superconnected, the relevance of reputation of more distant people and brands became more and more relevant. In fact it became so relevant that we had to invent search engines to handle that information overload.
-
As people quickly learned to cheat in the reputation game on search engines by stuffing repeated words into their webpages to influence their ranking in the search results, two gentlemen in Mountain View, California came up with an idea to compress reputation information and relevance in a unique way. Essentially the Google Page Rank algorithm counted references to web pages from across the web to better estimate its relevance. That meant that your reputation online became the compressed aggregate of information available about you on different webpages weighted by how much people give credit to it by linking to it.
This last phase lasted for 30 years from the 1990s until November 2022 when ChatGPT was introduced to the world.
Is AI the next generation of search engines?
Large Language Models (LLMs), the underlying technology behind ChatGPT, are able to answer free form questions in natural language in ways that mimic an expert or an individual educated in the field. Those answers come in seconds and that feels like a giant leap forward compared to the effort it takes to find answers from search engines.
Instead of browsing through a set of links to articles about the Best 5 mobile phones to consider and Top 10 laptops for Designers we can now just ask AIs the same question and get an immediate response.
Essentially LLMs allow us to capture statistical signals in the language and in that sense offer a peak into aggregate opinions over the corpus of human knowledge. Those snapshots of the world can be interrogated with natural language which makes them intriguing for researching any specific niche you might care about.
It is highly likely that this new technology that allows querying of unstructured snapshots of the world using free form human language is going to replace classical search engines. The improvements in both usability, speed and applicability of the responses alone justifies replacing all classical search engines with these new interfaces and the actions of major tech players demonstrate that this is already happening.
How do you track reputation according to AI?
As the new revolution of search woke up both the giants and the opportunists the landscape quickly became crowded with AIs that try to improve on the search experience. In fact there are so many, that trying to keep track of AI answers manually is just impractical.
If you're a restaurant owner in Vancouver Canada, you might check ChatGPT for Best 5 restaurants in Vancouver once or twice, but having to do it on a growing list of competing platforms (Microsoft Bing, Google Bard, Perplexity, ChatGPT, Claude etc.) or trying to see how your position changes in time in the recommendations provided by AI is just too much of a hassle.
In effect until very recently tracking reputation according to AI has only been possible manually.
What are the top tools for tracking reputation according to AI?
The simple answer is that there isn't much out there yet.
- For tracking information about individuals and their reputation in AI answers there was Have I Been Encoded.
- For tracking information about brands and their reputation in AI answers there was SEO for AI.
- For tracking information about recommendations and top lists according to AI there is Track AI Answers.
Full disclosure, all these are different projects built by the author of this article because at the time he could not find any other tools for monitoring brands, people and recommendations in AI outputs. All the three solutions have now converged under a brand Track AI Answers that you can check out in a quick video here:
That's all good for monitoring and tracking, but ...
How do you actually do AI reputation management?
There's plenty of articles out there teaching you how to track comments about your brand in social networks (classical reputation management) and how to use AI tools to either automate or improve responses to public conversations, but there seem to be no tools available yet to specifically focus on your reputation in AI responses (AI reputation management).
For now your focus should be on SEO (Search Engine Optimization) as the newest breed of AI models tend NOT to use the foundational model for responses but rather use the AI models to just interpret internet search results.
Your best chance to influence the answers would be to make sure the content in the search result top reflects what you want the AIs to say. Monitor the AI recommendations for your product category on Track AI Answers, do the SEO work to get better content to the search top and one day you will receive that happy email letting you know your position in the top list by AI has improved!
What is next for tracking your reputation on AI?
"Companies must monitor individual's comments regarding service 24/7."
Luckily we're not quite there yet with keeping an eye on reputation in AI answers. The mentioned services for example monitor AI responses automatically every two weeks, but you should expect the timeframe between new information becoming available online and it being reflected in AI answers to rapidly decrease in time.
for now To secure your brand in AI responses for the future of search means to prepare your monitoring tools to detect changes today!
