We had 15 mins scheduled for this meeting. A conversation between the Product Manager, Director of Customer Success, and me. Early on in the conversation, I posed a polarizing question.
“Why is no one stating the obvious? The number of intermediaries through which we grab the attention of the audience has increased significantly. Social media, WhatsApp, YouTube, Private Communities, Local Listings Platforms, Review Platforms. And now ChatGPT, Claude, Gemini. The list is endless and ever-increasing. To tackle this, businesses obviously need to monitor all these platforms. Yet, why is no one stating this up front? There needs to be an increase in marketing spend. “
Both the P.M. and Director shared their views on the question. Their answers got me thinking.
Business executives (and people in general) do not like to hear the truth upfront.
Saying, “to gain X, you spend Y”, straight up, no sugarcoating, does not land well with the audience. At least if your audience is a CEO or CFO who measures everything by the balance sheet. Being upfront and direct may not be the ideal approach.
Business executives look at marketing and operations as cost functions
Reduce costs as much as possible. Go to any local fruit seller, and they can explain why it is important to keep costs down. It helps keep the bottom line in check.
Take, for example, the business of running a hotel.
Few years back, the hotel had to deal with a limited number of marketing channels. A few major online travel agencies, corporate partners, and a focus on operations.
As of 2026, hotels now have to pay heed to influencers. What these influencers are suggesting about them in their private communities. The hotel needs to respond to every review in the ever-growing review ecosystem. They need to create video content on YouTube, Instagram, Facebook, and TikTok. As you can very well understand, the list is ever-growing.
With the help of AI, one can create yet another platform over a weekend. If the platform is any good, potential travel guests start utilizing it. It becomes yet another platform on which the hotel needs to be present, monitor, and create more content. Another new platform to focus on. The number of intermediaries just keeps on increasing.
This is as true for the hotel business as it is for any other category of business that has a service component to it. Be it automobiles, electronics goods, or financial services.
LLMs and their habit of hallucination
An LLM (or a Large Language Model) reads natural human conversational language and breaks down questions into smaller chunks. The technical term for breaking natural language into machine-readable chunks is known as query fan-out. Fan-out queries are searched in various databases. If not found, search takes place on the web before an answer is generated.
I agree, this is an extreme simplification. Yet at its core, this is the mechanism used by LLMs. Along with a whole host of statistical, probabilistic, and mathematical models before it shows the user any answer.
In this mechanism of retrieving and augmenting information before producing an output, LLMs try to reduce hallucination. Simply put, to generate a convincing answer, it can make up facts and statements that may be completely false. That is what hallucination is.
To avoid hallucination, LLMs rely heavily on trust signals. These trust signals are accumulated across various sources, which helps in grounding the information before an answer is produced.
It’s like me asking 20 individuals who have visited North Korea for their opinion. “Is it safe to visit North Korea with my wife?”
If majority of responses are positive, I can ground myself with real-world experiences from previous travelers before making my travel plans.
If I am not getting a clear positive answer, I may refrain and choose another destination for my vacation.
That’s me grounding information from multiple sources.
LLMs do a similar thing; the only difference is the scale at which they are performing grounding.
As an individual, I can ask a maximum of 20 of my friends who have visited North Korea. When an LLM is asked a question, they look at 100s, if not 1000s of data points to ground the answer. Their ultimate motive is to ensure answers produced are not hallucinated.
The product manager (an alumnus of IIT Roorkee), S.J., suggested: “Businesses should create, curate, and manage content if they want to ensure an increase in AI search visibility”. He further added, “The content from various platforms needs to be consistent, coherent, and trustworthy.”
But all this monitoring, managing, and creating content on different digital platforms requires more time, effort, and money. Suddenly, cost optimization starts to look a lot more daunting in real life than on the balance sheet.
What used to work a few years back is not applicable anymore
In some cases, it’s a total 180-degree flip.
“Increasing AI search visibility is a goal that cannot be achieved by one department alone in a silo”, said S.J.
“Your marketing can be top-notch, but with bad customer experience, your reviews may tell a different story. LLMs would get confused with this discrepancy”, concluded S.J.
It may not recommend your business simply because it doesn’t have enough proof to recommend you vs your competitors.
A few decades earlier, there were a limited number of channels through which marketing was performed.
For a hotel business, it was limited to travel partners, Trade Shows, Phone calls, etc. With the introduction of the internet, a whole host of new platforms emerged. From search marketing to OTA (Online travel agencies) to social media.
We are noticing something similar with the introduction of AI. These sophisticated models use 1000s of data points across 100s of channels that are scraped as grounding knowledge before the answer is produced.
These new data sources may not seem like a dedicated channel in themselves. Nonetheless, they are being used by AI before recommending your business when a user asks a question.
Let’s look at an example for a hotel business
A couple who stayed at your property had a very bad experience. They did not bother to write a bad review on Google or TripAdvisor. But they run a small travel blog (a personal website of sorts). A detailed account of their experience of staying in the hotel gets written on the blog. AIs may utilize this little-known blog as one of the sources of information grounding.
Similar is the case when it comes to other platforms or sources.
Users share opinion with their friends and family about a hotel and their bad experience in WhatsApp and Facebook Groups. There is almost no way for you (as a business) to monitor such conversations. But those might well be used by the LLM to create a digital persona of your hotel.
There are 2 ways of tackling such cases.
One of them is ensuring you have great operating procedures that lead to a superior customer experience. That way, you try to influence users right when they are interacting with your business. By delivering an exceptional customer experience. If this happens, the chances of users badmouthing you get eliminated at the bud.
The other is to create more content (content created by the business and 3rd party sources).
For both use cases, as you have noticed, as a business you would need to shell out more dollars.
In the first case, money goes to deliver superior customer experience. Be it in terms of manpower or equipment.
In the second scenario, more content (content can be multimodal media, i.e. text, audio, video, augmented/virtual reality) gets created. Spending has to increase to maintain consistency and distribution of such content.
Accounting philosophy is not going to change anytime soon
How the CFO creates and looks at the balance sheet is something I don’t think will change in the very near future. Yet, there’s no hiding from the fact that expenses of running a business will only increase as we march towards a world where AI is integrated in each and every activity we perform.
But what happens when you go to the CEO/CFO asking for more resources in this age of AI? You are stopped and told to use AI instead of increasing human capital investment.
Fair argument, but then again, let’s consider the downsides of using AI blindly. Especially from a marketing perspective.
AI tools can help us identify the type of content to create. It can also automate the content creation mechanism. But you still need a human to vet the content. Is the content produced adding any value for the end user?
How will this content be any different from the content being produced by the business next door? They may use the same tool.
The differentiating factor can be an actual human. Since no 2 human beings in the world are a carbon copy of each other. Each will have their own unique experiences. The manner in which the tool is used is where the differentiation happens.
Tools can be mass-produced at a very low price; the experiences and unique POV of human expertise require a lot of investment. Investment of time, effort, and money. Wherever more money investment is coming in the picture getting a buy in becomes challenging. So what next?
Operations to the rescue
I have a contrarian view on how businesses should tackle the concept of increasing AI visibility. It starts with operations and not marketing. When the user of the product or service becomes happy, inevitably they would share their happiness with others in their circles. Some of it may be written on different digital platforms; some of the praise would be propagated through word of mouth.
Although capturing this feeling of customers being happy using your product and service can be extremely difficult to measure, yet, that becomes the only success for measuring success. If a customer decides to transact with your business again and again, that is a unique data point signifying operational happiness and superior customer experience. Customer is happy, which is why repeat business is rising. Simple data that can be measured by anyone is any organization.
Focus on the customers you already have
Increasing marketing and operational budgets might be very difficult to get by from the CEO or CFO. But as businesses, we still need to start somewhere. The philosophy I put out there for gaining customers starts by keeping the customers that are already with you, HAPPY.
Trying to see what more can be done for these customers? How can the customer become advocates, extended salespeople residing outside the organization?
Increasing operational investments can actually put us in the driver’s seat when it comes to more visibility in AI searches. Superior customer experience leads to customer advocacy and word-of-mouth marketing.
In this age where marketing materials can be pumped out at breakneck speed, and audiences are inundated with more content than ever, the only differentiating factor may come down to operations.
Am I happy using your product and service, and would I use it again? If the answer is yes, search visibility will eventually increase. Albeit a little slower than what your sales team would like it to be.
This AI visibility gain based on strong operational foundations would be very difficult to displace. Even if a competitor enters the market with deep marketing pockets. It is not outlandish marketing but superior customer experience delivered through operations, creating your business’ moat.
Till next time,
DB
27th September 2026