The Do’s & Don’ts of Building a Cash-Positive Company in One Year

Key Insights from Artem Gladkikh, CEO and Founder of Signum.ai, a New York-based automated B2B Demand Generation AI-powered startup.

How did you start your entrepreneurial journey?

My first project was about identifying clothes in photos and searching for similar ones on various marketplaces. Once I sold that platform, I searched for new opportunities to create other useful products. I’ve always been interested in playing around with data, particularly since the number of data is growing so rapidly and analyzing data has become increasingly challenging. The issue of big data analysis has never been so important, so I was inspired to go in this direction.

And so I started the company called Test4startup. We analyzed various trends looking at how they fit the market. For example, if you wanted to build an Uber for kittens, you could check whether or not the market is ready for that idea. We engaged data from different VCs such as Crunchbase, and so it was possible to track growth in the number of deals occurring in the space.

However, we realized that we wouldn’t be able to make money with this product. This is when Signum.ai emerged. All of our ideas and groundwork were quite useful, and we started to offer more substantial services apart from selling analytics and forecasts. We dug deeper to collect leads and parse audiences, which eventually improved the marketing campaigns’ performance.

How did you manage to keep a healthy monthly revenue growth rate — +30%?

At the start, we moved in a chaotic way, which is normal for any company at the early stages when the ambiguity is high. One of our team’s strengths is the ability to test hypotheses fast and on a small budget, and grasp what works and what doesn’t quickly. We devote the time and number of tasks to testing hypotheses, reacting instantly, researching, and implementing new ideas. So, building a model that utilizes cheap, responsive, and quality tests is the key to success in our experience.

At some point along our journey, we realized that we had to choose one direction. Even if there was chaotic movement, we’d be at least moving forward along one path. And so we chose to focus on developing a product that addresses the vital issues, from driving revenue, to generating leads and getting new clients.

We tested a large number of hypotheses working on customer development. Every single day we would test something new. We generated new ideas and offered them to clients straight away. Generally, one out of thousand offers would win, and from there we would run the first pilot and build out the relationship with the client. We were building custdev by crude methods with a tangible outcome. From the very beginning, we saw the US market as a priority and focused our energies there.

In what way was Starta Accelerator beneficial for you?

To enter the US market, it’s critical to understand how it works. We started by attending conferences, and in this way the accelerator program was very helpful. Once we joined, we were able to test various hypotheses. We also managed to meet clients in person and offer our product directly, which is so helpful when building relationships. This is when we realized that the ideas we had generated in Moscow didn’t necessarily translate well to our target clients in the US and around the world.

Once we better understood the direction we wanted to go, we started to grow rapidly. Joining the accelerator helped us think globally, and achieve a product-market fit. It is essential to work out these details at the earliest stages because failing to think globally at the outset almost certainly means going back to the drawing board in the future.

What does one need to test hypotheses effectively?

Our first rule is to always listen to the client. Whatever they say becomes the primary goal. This is a major pitfall among founders; they think they know the market and how it works, so when clients ask for something else, they think: “You don’t understand a thing. I know the market, let me explain how it should be.” When people tell you directly: “Do this, and we will buy,” listen! Even if their thoughts don’t fit your worldview, we have found it is most effective to follow their lead.

The second component is testing quickly and cheaply. Our model is to run one cycle per week. Experiment, get results, adjust, and repeat. For example, you might start a cycle, test it, get the outcome, generate an ad creative, start a Facebook campaign, and get five demo requests. If you’re not satisfied, you generate new creatives, run the campaign, and get 50 requests. Sounds better, doesn’t it? Then figure out why it worked by getting in touch with clients and collecting their opinions. This is the way to build and polish new products. Our guiding principle is to stay hungry.

What impact has COVID-19 had on the company?

COVID-19 has sped up our growth trajectory at Signum.ai. Before the pandemic, we had around two requests per day, which has since turned into 40. Our main task in the midst of this escalation has been to ensure the stable onboarding of new clients and not lose quality.

As for the pre-seed round, COVID-19 has changed the rules. Everything took longer. We didn’t have any difficulties with fundraising — about 20 investors were ready to join the round. We chose based on several criteria: first, it had to be smart money; second, we were interested in professional investors only; and finally, we wanted to engage solid funds. We were fortunate — the investors we sought agreed to join, so we stopped pursuing alternatives.

What is your advice to founders when raising funds?

Always think about the business first. If attracting investors becomes your priority, it will negatively impact the business. Whereas if you aim to build an excellent business, trust that the right investors will want to be a part of it and join the round. There are two business models: 1) running a business to make money; and 2) running a business to please the investors. In the case of the latter, the cover is usually brilliant, but at the end of the day, nothing works. I believe everyone in the market understands who is who; either your startup develops a legitimately cool product, or it fizzles out after lots of hype on social media.

Here is my advice: don’t think about attracting investors. When you get to the point that there is not enough money to scale, just bring the calculations to investors and make a promise: this is the sum they invest, and this is the outcome they’ll get. This will be compelling to them.

What challenges have you faced along the way?

I’ve faced two issues. First, a lack of experience as a professional top manager. Yes, I have acquired management skills from building small teams of up to 10 people. But before Signum.ai, I hadn’t gotten any experience as an executive who would be managing the team of managers.

The other challenge is rapid growth. When the growth is too intense, it leads to a situation where it’s impossible to manage it all, and so I had to employ new people at lightning speed. This includes onboarding and training, which is time consuming when done right. As the inflow of leads increased, we started replying to clients slower, and naturally, the level of general dissatisfaction rose. This problem is not easy to resolve, and as a result, our performance got worse. There was a snowball effect: you can’t focus on everything simultaneously, and so you jump from one task to another, and eventually don’t complete any tasks at a proper level.

As the company keeps growing, the decision-making process and number of tasks get more complicated. We try to think of management as a “domino” effect. When you push one tile, that will cause a certain result in 10 steps. It takes time to get there, but it’s important to anticipate needs and complete the tasks now that will trigger those steps. We’re always making decisions in real time, thinking several steps ahead about what we’d like to accomplish. It gets better with time. When the delegation system is established, you should have only two processes to focus on: developing strategy and managing the team.

What are the conclusions you drew out of these challenges?

One key insight is that it’s not only you who is doing a great job. There are many people that are better at things than you are. So even if their vision does differ from yours, it’s important to consider their informed stance, because they might be right.

Also, it’s vital to control stress levels. Panicking is rarely useful, so it’s often better to go with the flow and learn from your mistakes, no matter how catastrophic. For example, if you’re experiencing overwhelming, rapid growth, at some point you get used to it and learn how to swim in the deep end. You start seeing hiring and firing 100 people as a normal process. I can’t say that I feel comfortable with it yet. I believe if you’re building a company, growing rapidly, and feeling comfortable — there’s something wrong. Even though the decision-making process simplifies over time to some extent, the level of stress remains more or less the same. It’s not possible to found a company and then fly to Bali straight away, grab a smoothie, and enjoy life while everything maintains itself and grows on its own. Unfortunately, startups don’t work like that.

What are your plans for developing the company?

We plan to stick to the same growth rates. It’s easy to grow when you’re small. The rule of small numbers means that when you generate $500 more this month than in the last month, you rock. It’s much more difficult to grow when it comes to hundreds of thousands of dollars. Growth can be vertical or horizontal. If horizontal, simply come up with new ways of selling existing products more. As for vertical growth, you have to reimagine the approach to sales and product in a global sense.

In our case, we’re working on horizontal growth right now. We employ new specialists to enlarge the marketing team. We also work on LTV with those who can make more out of one client. Generally, we have quite friendly relationships with clients. That’s essential to ensure each client gets high-quality services, and why cross-sales are easy to do for us. The difficulty for us right now is that we don’t have enough people.

Our primary goal right now is to build a system that works, no matter who is managing various processes. At the next level (round A), you don’t get investments just for your vision or charisma. What investors are looking for is a functional, sustainable system. In our case, we tested metrics like retention and unit economics to demonstrate our functionality.

The other important task is developing your own technology solutions. Our current Signum.ai products are based on GPT-2, which helped us learn to generate high-converting content for marketing campaigns. Recently we got access to GPT-3, which is at a much higher level. Compared to GPT-2, the number of parameters involved grew 100 times: from 1.5 to 175 billion. This means we can generate content of excellent quality. At the moment, we’re on a fully-booked waiting list for access, and the OpenAI team is overloaded with tasks. It’s an incredible competitive advantage for us, as there are no other products that have this technology in their toolkit on the market right now.

How do you differ from your competitors?

We have strong technology and sales teams that grow certain metrics, and their teamwork is very efficient. When it comes to the CIS market, there are no marketing and sales agencies that haven’t got in touch with us at some point. We have high and unyielding standards, so it’s not easy to get onto our team. No matter how charming a person is, even if they’re a close friend or relative, it is crucial to separate personal and business relationships. At our company, personnel demonstrate excellent performance and solid outcomes. We’re here to build a number one company, and we are focused on results.