Between them, the two co-founders have closed several businesses. Marketing Kernal, their new venture in AI visibility and sales software, is an experiment in whether founders can actually learn from failure rather than just talk about it.
Entrepreneurs like to say that failure is a teacher. Fewer of them can tell you exactly what it taught.
Vaibhav Palhade and Pratyush Shrivastava can. When the two began working together, they did something most founders avoid. Before discussing what they wanted to build, they wrote down why their previous companies had failed.
The list was not short. Palhade, an engineer from Akola in western India who built an earlier career as a screenwriter and novelist, counts three or four earlier ventures that did not survive. Shrivastava, a technologist from Jaipur, ran a startup that also closed.
“We did not start with a pitch,” Palhade says. “We started with a post-mortem. Then we asked a harder question: if we are not careful, will we make the same mistakes again? The honest answer was yes.”
That exercise, followed by roughly two to three years of research, eventually produced Marketing Kernal, formally established in 2026. The company works on AI visibility, which concerns how businesses appear in answers generated by tools such as ChatGPT and Gemini, and is developing software for lead generation and sales conversations. Here is what the founders say the list contained, and how it changed what they built.
Mistake one: building before anyone asked
Shrivastava’s first company did not fail because the technology broke. “The product worked,” he says. “We had just built a lot of it before anyone told us they needed it. Good engineering and a good business are not the same thing. I learned that expensively.”
At Marketing Kernal, the order is reversed. Ideas are tested against real business problems before significant engineering time is spent.
Mistake two: mistaking volume for progress
Palhade spent years in search and digital marketing, running growth programs for software companies selling overseas. Several of his earlier ventures, he says, made the same error as many marketing departments. They optimized for more: more traffic, more leads, more activity.
“Volume feels like progress because it fills dashboards,” he says. “But if the leads are wrong, you are just failing at a larger scale.”
That thinking shaped Outreach Kernal, the company’s first product, a B2B prospecting tool that the founders say is built around verified data and careful buyer research rather than mass messaging. A beta is scheduled for November.
Mistake three: optimizing for yesterday’s channel
In his earlier companies, Palhade says, he kept optimizing for the channel he already understood. This time the founders asked where buyers would be in three years.
Their answer was AI assistants. In February 2024, Gartner predicted that traditional search engine volume would fall 25% by 2026 as AI chatbots and virtual agents took a larger share of queries. Whatever the final number, the direction has changed how businesses get discovered.
Other products in development, which the founders are keeping quiet for now, are aimed squarely at that shift, and at the sales conversations that follow it.
Mistake four: trying to do everything
“In my earlier attempts, I tried to be good at everything,” Palhade says. “That is how you end up average at the things that matter most.”
This time the division is strict. Palhade leads marketing, product and go-to-market. Shrivastava leads technology and infrastructure. They disagree often, they say, and they consider that a feature.
Shrivastava’s summary is characteristically short. “One of us has to be the optimist,” he says. “It is never me.”
Mistake five: promising what you cannot control
The AI visibility field has filled up with bold claims, including guaranteed placements in chatbot answers. Marketing Kernal makes none, and the founders connect that directly to their history.
“When you have failed before, you become allergic to overpromising,” Palhade says. “The model decides what it says. We control the method, the data and how honestly we report results. That is the whole promise.”
Mistake six: treating rules as an afterthought
The company is building for businesses in the United States, the United Kingdom, the UAE, Canada, Australia and India. Some of those markets, particularly in Europe, impose strict data and AI rules.
“If you design for compliance at the start, it costs a little,” Shrivastava says. “If you design for it later, it can cost the company.”
Mistake seven: hiding the failures
Perhaps the least common lesson is the one the founders practise most visibly. Many entrepreneurs scrub old ventures from their biographies. Palhade and Shrivastava talk about theirs readily, without naming the companies but without disguising what happened either.
“Investors, customers and future colleagues are going to find out anyway,” Palhade says. “I would rather they hear it from me, along with what changed. A founder who has never failed has not been tested yet. A founder who has failed and learned nothing is a risk. We are trying to be neither.”
The quiet years
For Palhade, the venture is also a return to public life. His most recent novel, Echoes of Yesterday, was published in 2024, and he released nothing for close to two years while the research continued.
“People assumed I had stopped,” he says. “I had only stopped publishing.”
Neither founder offers predictions about where the company will be in five years. Both say that is part of the lesson. What they do offer is a written record of what went wrong before, and a company designed, line by line, not to repeat it.
“Failure is only useful if you are honest about it,” Shrivastava says. “We wrote ours down and read it back. Most of the company came from that.”