2. Leads History - B2B

As mentioned in the 
1. Leads History - B2C
, the shift to B2B started when we went around to events trying to sell our product to people, and they started answering that they wouldn’t personally buy it but that they knew of a company that could be willing to pay for a solution for their problems. Many of the possible use cases discovered during the event wouldn’t be plausible with ChatGPT.

We then started asking companies and startups if they were interested in our product. For a time, we focused on crypto companies, but the lead turned out to not be as good because they were mostly interested in using AI as publicity for getting VC money and not in using our solutions for real-world applications. This mindset transformed our solution into a “vitamin” instead of a “painkiller.”

More or less at the same time that we were figuring that crypto companies weren’t a good lead, we joined the Techstars program. During the Mentor Madness month, we started validating what we already suspected: B2B was the way to go since companies in the Web2 world were interested in what we were doing but not crypto and Web3.

Then, we began searching for different markets and interviewing people using the Techstars contacts. We ended up finding that companies in the insurance and banking industries might be interested in our solution due to privacy concerns related to centralized AI solutions. Those leads regarding privacy didn’t lead anywhere because these companies required a lot of validation points and certificates that a startup doesn’t have. Besides, the sales cycles took years, and we just had 1 year of runaway at the time.

With this information, we decided to abandon the insurance and banking industries in the short term and think more about other types of industries that would be more desperate to use our solution. This made us talk with more people, which in turn led us to the market we are currently pursuing:

The story began with one of our family members calling us because he knew that we worked at an AI startup, and his company needed help in that regard. The company builds machines for the plastics industry and offers two years of free maintenance for the machines that they sell, and they see this as one of the competitive advantages they have. However, they recently started to export to another continent, and problems that previously took them a couple of days to fix now took them weeks, besides the traveling costs. Due to that, they started to try to find out ways to reduce the costs, and they discovered that a lot of the problems that they usually fixed had already occurred years before. However, everyone had forgotten about them because they were stored in PDF reports. So, they wanted a “ChatGPT for their business”. An AI model that would have all the knowledge of the company, including reports, clients’ data, and accounting documents. That model should then be capable of doing graphs, monthly reports, and answering any kind of questions related to the company’s data so they can become more productive and make better decisions. They asked consultant companies to do this tool, but the consultant companies were sending budgets of 3 years of development and around $1M. This was mainly because the consultant companies were thinking of building a back office first to digitalize all the information using forms, changing the company processes, and so on. This was way out of the comfort zone for the manufacturing company, as they are a small company with employees that have worked in the same way for more than 20 years. Changing the processes or paying $1M was out of the question. So they contacted us to find out if the prices were correct, and by explaining their problem, we realized that we could actually build a solution for this without the back office, using only AI. This would lower the development time to 8 months and the development costs to $75k. We made them this offer, and they are currently waiting to receive more projects abroad to sign a contract with us.

The problem of this company has sparked a cascade of ideas on our side because the above problem isn’t exclusive to that company. The three co-founders worked with or within consultant companies in the logistics, sports, banking, education, and telecom sectors, and by looking at all the projects we had developed, they all followed similar patterns: companies actually wanted to make better decisions based on the data they were gathering, but for that, they were using old fashion software when this is too expensive and the worst solution in most cases in the AI era.

We then started contacting everyone we know related to industries, including old colleagues, and one of them is an old acquaintance who is now trying to sell our product in the metropolitan area of Lisbon. We negotiated with him a 10% commission on any deal that we would do with a business that he would introduce to us. He has contacts in multiple sectors besides manufacturing, like banking, GovTech, and insurance. He is currently trying to get us a deal with an insurance company related to using AI to handle the company’s data, and he has also been leading the way in using our system to obtain useful statistics about public students (GovTech).

Parallel to that, we started making Lean Canvas and B2B and B2C interviews using the JTBD framework with the help of one of our advisors (Alex Lumley). The main goal was to confirm if B2B was indeed the right lead and, if so, which industry would be the best to start. While doing that, we got an idea for a potential pivot in case this lead ends up in a dead end. The idea was automating the JTBD framework using AI in order to help businesses or political parties to understand better why surveys and polls have certain results. Part of our team has been testing this idea mainly on free time in a political party in Portugal, with really satisfactory results, reaching non-obvious conclusions and different from political analysts' guesses.

We also used our contacts from Techstars to interview a salesman from Ireland who worked in manufacturing and worked with Intel for 11 years. He agreed with our vision and believed there was a market for our product. However, he suggested that we go after the real estate sector instead of the manufacturing one. He believes there is also a market for manufacturing, but it is only more tightly regulated. He suggested that we pursue the sales by using a salesman like he was doing instead of going door to door, with the salesman receiving a 10% commission if things worked out. He said he could talk to a few companies that might be interested in our solution.

After the agent strategy was recommended to us, we started looking for more salesmen. While we were asking around, a familiar said he knew a Portuguese politician who sells defibrillators in the healthcare industry. As he has good connections, he could potentially introduce us to a lot of people. We went to talk to him, mainly hoping that he could help us sell our product in the manufacturing, healthcare (as an AI system for hospital management, for example), and GovTech industries or that he could give us his honest opinion regarding our political lead. Regarding manufacturing, he isn’t very enthusiastic but said he could give us two different contacts. In Healthcare, he is excited about the idea, especially if we would like to do new drugs using AI, but warned us about legal restrictions and that it could be a crowded space. About GovTech, he is really excited and wants to present to us some people related to the defense industry, including maybe the Portuguese Ministry of Defense. Lastly, regarding politics, he was really passionate about the idea and found it to have big possibilities, but most of the political parties in Portugal wouldn't have the money to pay for our services.

Some extra details about each one of these people can be found on the 
2.3. Lead Subsections
 and the 
2.2. Why B2B wants us
 pages. From the leads mentioned, we made a list of what we believe are our priorities and how we divide them here: 
2.1. Priorities in Leads
. A summary of the main insight we have about the interviews we are conducting can be seen here: 
2.4. Insights
.

2. Leads History - B2B

As mentioned in the 
1. Leads History - B2C
, the shift to B2B started when we went around to events trying to sell our product to people, and they started answering that they wouldn’t personally buy it but that they knew of a company that could be willing to pay for a solution for their problems. Many of the possible use cases discovered during the event wouldn’t be plausible with ChatGPT.

We then started asking companies and startups if they were interested in our product. For a time, we focused on crypto companies, but the lead turned out to not be as good because they were mostly interested in using AI as publicity for getting VC money and not in using our solutions for real-world applications. This mindset transformed our solution into a “vitamin” instead of a “painkiller.”

More or less at the same time that we were figuring that crypto companies weren’t a good lead, we joined the Techstars program. During the Mentor Madness month, we started validating what we already suspected: B2B was the way to go since companies in the Web2 world were interested in what we were doing but not crypto and Web3.

Then, we began searching for different markets and interviewing people using the Techstars contacts. We ended up finding that companies in the insurance and banking industries might be interested in our solution due to privacy concerns related to centralized AI solutions. Those leads regarding privacy didn’t lead anywhere because these companies required a lot of validation points and certificates that a startup doesn’t have. Besides, the sales cycles took years, and we just had 1 year of runaway at the time.

With this information, we decided to abandon the insurance and banking industries in the short term and think more about other types of industries that would be more desperate to use our solution. This made us talk with more people, which in turn led us to the market we are currently pursuing:

The story began with one of our family members calling us because he knew that we worked at an AI startup, and his company needed help in that regard. The company builds machines for the plastics industry and offers two years of free maintenance for the machines that they sell, and they see this as one of the competitive advantages they have. However, they recently started to export to another continent, and problems that previously took them a couple of days to fix now took them weeks, besides the traveling costs. Due to that, they started to try to find out ways to reduce the costs, and they discovered that a lot of the problems that they usually fixed had already occurred years before. However, everyone had forgotten about them because they were stored in PDF reports. So, they wanted a “ChatGPT for their business”. An AI model that would have all the knowledge of the company, including reports, clients’ data, and accounting documents. That model should then be capable of doing graphs, monthly reports, and answering any kind of questions related to the company’s data so they can become more productive and make better decisions. They asked consultant companies to do this tool, but the consultant companies were sending budgets of 3 years of development and around $1M. This was mainly because the consultant companies were thinking of building a back office first to digitalize all the information using forms, changing the company processes, and so on. This was way out of the comfort zone for the manufacturing company, as they are a small company with employees that have worked in the same way for more than 20 years. Changing the processes or paying $1M was out of the question. So they contacted us to find out if the prices were correct, and by explaining their problem, we realized that we could actually build a solution for this without the back office, using only AI. This would lower the development time to 8 months and the development costs to $75k. We made them this offer, and they are currently waiting to receive more projects abroad to sign a contract with us.

The problem of this company has sparked a cascade of ideas on our side because the above problem isn’t exclusive to that company. The three co-founders worked with or within consultant companies in the logistics, sports, banking, education, and telecom sectors, and by looking at all the projects we had developed, they all followed similar patterns: companies actually wanted to make better decisions based on the data they were gathering, but for that, they were using old fashion software when this is too expensive and the worst solution in most cases in the AI era.

We then started contacting everyone we know related to industries, including old colleagues, and one of them is an old acquaintance who is now trying to sell our product in the metropolitan area of Lisbon. We negotiated with him a 10% commission on any deal that we would do with a business that he would introduce to us. He has contacts in multiple sectors besides manufacturing, like banking, GovTech, and insurance. He is currently trying to get us a deal with an insurance company related to using AI to handle the company’s data, and he has also been leading the way in using our system to obtain useful statistics about public students (GovTech).

Parallel to that, we started making Lean Canvas and B2B and B2C interviews using the JTBD framework with the help of one of our advisors (Alex Lumley). The main goal was to confirm if B2B was indeed the right lead and, if so, which industry would be the best to start. While doing that, we got an idea for a potential pivot in case this lead ends up in a dead end. The idea was automating the JTBD framework using AI in order to help businesses or political parties to understand better why surveys and polls have certain results. Part of our team has been testing this idea mainly on free time in a political party in Portugal, with really satisfactory results, reaching non-obvious conclusions and different from political analysts' guesses.

We also used our contacts from Techstars to interview a salesman from Ireland who worked in manufacturing and worked with Intel for 11 years. He agreed with our vision and believed there was a market for our product. However, he suggested that we go after the real estate sector instead of the manufacturing one. He believes there is also a market for manufacturing, but it is only more tightly regulated. He suggested that we pursue the sales by using a salesman like he was doing instead of going door to door, with the salesman receiving a 10% commission if things worked out. He said he could talk to a few companies that might be interested in our solution.

After the agent strategy was recommended to us, we started looking for more salesmen. While we were asking around, a familiar said he knew a Portuguese politician who sells defibrillators in the healthcare industry. As he has good connections, he could potentially introduce us to a lot of people. We went to talk to him, mainly hoping that he could help us sell our product in the manufacturing, healthcare (as an AI system for hospital management, for example), and GovTech industries or that he could give us his honest opinion regarding our political lead. Regarding manufacturing, he isn’t very enthusiastic but said he could give us two different contacts. In Healthcare, he is excited about the idea, especially if we would like to do new drugs using AI, but warned us about legal restrictions and that it could be a crowded space. About GovTech, he is really excited and wants to present to us some people related to the defense industry, including maybe the Portuguese Ministry of Defense. Lastly, regarding politics, he was really passionate about the idea and found it to have big possibilities, but most of the political parties in Portugal wouldn't have the money to pay for our services.

Some extra details about each one of these people can be found on the 
2.3. Lead Subsections
 and the 
2.2. Why B2B wants us
 pages. From the leads mentioned, we made a list of what we believe are our priorities and how we divide them here: 
2.1. Priorities in Leads
. A summary of the main insight we have about the interviews we are conducting can be seen here: 
2.4. Insights
.