3 weeks ago
14
Golf has always been perceived as an expensive and exclusive sport, often seen as a pursuit of the elite and a marker of social status.
Tushar Narang, an avid golfer and founder and CTO at Golfoy.com, thinks otherwise.
In an exclusive interview with AIM, he debunked several myths about the affordability and accessibility of the game based on his personal experience.
In fact, drawing comparisons to more popular recreational activities, he explained, “Going to a mall for a bowling game costs you ₹1,500 for less than 40 minutes.”
He further mentioned that a golf session typically costs ₹1,500, allowing one to play all 18 holes and spend several hours at the club.
At its core, the barrier to entry lies more in perception than reality, Narang argued. For him, the mission is clear. “Our aim as golfers is to bring more people into golf, make golf accessible, make it approachable, and ensure they can try the right equipment.”
Delving deeper into the importance of having the right equipment to truly enjoy golf, he also highlighted how AI plays a key role in ensuring players have gear that suits them perfectly.
Golf and AI
Narang pointed out that the blend of AI, sensors, and simulation reshapes how the game is played and experienced.
Comparing the sport with cricket, where using a bat that’s too heavy makes it harder to play effectively, Narang said using the right golf equipment tailored to an individual’s needs is important. In golf, equipment weight and specifications impact performance to a greater extent.
Keeping this in mind, Golfoy developed a cutting-edge tool unlike anything currently available in the market—a smart equipment selection guide for golfers. This tool allows users to answer a series of simple, straightforward questions without requiring any advanced knowledge. Based on these basic inputs, the system provides recommendations that are up to 95% optimal for equipment suitability.
A beta version of the tool has been in use for about three years, during which a significant amount of data was collected. Based on that, around 80% of users reported that the recommended equipment was ideal and suited their needs.
Out of a total of 1,20,000 golfers in India, Narang noted that Golfoy has about 30,000-32,000 golfers registered on the website, with around 45,000 golfers visiting it on a monthly basis.
Now, as revealed by Narang, the company is working on an advanced version of the tool, which will be enhanced with a small sensor that can be attached to the equipment. The sensor will help users determine whether the equipment they currently use is ideal for them or if they should consider other options.
What about the Indian Market for Golf?
Narang highlighted that the golf market in India is still relatively small compared to the global scene.
He further explained that while Golfoy is a mature player in the Indian market, occupying 15-17% of the market, the situation is different globally as e-commerce in golf is a key focus area. As a result, it provides Golfoy the opportunity to focus on customer experience and other improvements.
Explaining the difference in product availability, Narang noted that while a particular product might be easily found at a local store in India, only about 20% of the products that Golfoy offers are available at golf courses.
“We deliver to one out of three Indian golf courses every month.”
In short, Golfoy is using its platform to address some global gaps in the market, specifically when it comes to what to buy and how to buy it.
AI Can Change Golf
“Because golf is a very fickle sport, the smallest differences give a massive result. You are hitting a ball which is 250 yards away. And even if you differ 1 mm, the result ahead will be 50 feet apart, ” Narang explained.
Given this level of precision, many golfers—especially professionals—rely heavily on data to improve their performance. AI can furnish that data, helping golfers know what degree or angle of attack to use.
“But there is no AI tool. There is nothing that advises you what you should buy,” he added. Notably, Golfoy has introduced the concept of club fitting in golf, a process that involves detailed testing and calibration.
“We have a machine that calibrates all the data. For every club, [we analyse] what results have come out. Then, we run permutations and combinations to ensure that what they have is the best product,” he explained.
By using AI, Golfoy is simplifying equipment recommendations. This method helps narrow down choices.
Narang emphasised that Golfoy’s goal is to bridge the gap between those who seek in-depth, offline fittings and those looking for quicker, AI-driven recommendations.
Why GenAI is not Working for Golf
Narang clarified that generative AI models are not applicable in their case. Instead, they rely on machine learning, which requires on-ground data.
“So we are using PHP, Laravel and JavaScript to actually build our AI.”
Notably, Golfoy’s team had previously experimented with OpenAI, particularly in customer interaction models. The company aims to integrate OpenAI with the equipment selection guide they are developing. The plan is for users to simply ask questions, with all queries routed through OpenAI. The responses and results will be processed through OpenAI and then passed to their LLM. Any data that needs to be presented to the customer will also be filtered through OpenAI before being shown.
However, he pointed out that OpenAI, currently, is quite broad-based and tends to veer off topic easily. “We are not able to control the narrative. We want it to only talk about golf. I have tried models when you are talking about golf, and it starts talking about life. So…it’s not very trustworthy right at this point,” he explained.
Even when using OpenAI in a controlled way, he explained, “We’ve got to feed it data and ask OpenAI to only go through that data and not anything else.”
He described the current approach as more traditional. “The only difference is that the person has to choose what they want to know—they need to select a different selection guide—and through that, they will have to answer a series of questions. After that, the data will be processed and the results will be delivered. It’s a more orthodox way of presenting the data to the customer.”
Narang shared his long-term ambition of developing a modern AI model for golf. “That is something we ultimately want to build. We might use existing AI models for that. But right now, we are building a product that just gives you the right equipment that you require.”
He further added his vision to develop an open-source model that can be used by a number of people around the globe. “We are bootstrapped since day one. We have never taken funds. We have been profitable since day one,” he concluded.
