
In 2018, I started an AI chatbot company...today, we have over 4000 paying customers and ChatGPT is changing EVERYTHING
Intro:
5 years ago, my co-founders and I ventured into the space of AI chatbots and started our first truly successful company. Never in a million years did I see myself in this business and we truly stumbled upon the opportunity by chance. Prior to that, we ran a successful lead generation business and questioned whether a simple ai chat product would increase our online conversions. Of the 3 co-founders, I was skeptical that it would, but the data was clear that we had something that really worked. We built a really simple MVP version of the product and gave it to some of our top lead buyers who saw even better conversion improvements on their own websites. In just a matter of weeks, a new business opportunity was born and a major pivot away from our lead generation business started.
Our growth story:
Startup growth is really interesting and in most cases, founders aren't really educated on what a typical growth curve looks like. While we hear about "hockey stick" growth curves, it's really atypical to actually see or experience this. From my experience, growth curves take place in a "stair curve". For example, you can scrap your way to a $100k run rate without much process or tracking. You can even get to $1 million ARR being super disorganized. As you start going beyond $1M ARR, things start to break and growth can flatten out while you put new processes and systems in place. Eventually you'll get to $2M or 3M with your new strategy and then things start breaking again. I've seen the process repeat itself and as you increase your ARR, the processes and systems become more difficult to work through...mainly because more people get involved and the product becomes more complex. When you do end up cracking the code in each step, the growth accelerates faster and faster before things start to break down and flatten out again.
Without getting too much into the numbers, here were some of our initial levers for growth:
- Our first "stair" step was to leverage our existing customer base from our prior lead generation business. Having prior business relationships and a proven track record made it really simple to have conversations with people who already trusted us to try something new that we had to offer.
- Stair #2 was to build out a partner channel. Since our chat product involved a web developer or agency installing the chat on client sites, we partnered with these developers and agencies to leverage their already existing customer bases. We essentially piggy-backed off of their relationships and gave them a cut of the revenue. We built an internal partner tracking portal which took 6+ months, but it was well worth it.
- Stair #3 was our most expensive step, biggest headache, but added the most revenue. After COVID, we had and SDR/Account Executive sales team of roughly 30 people. It added revenue fast, but the payback periods were 12+ months so we had to cut back on this strategy after exhausting our universe of clients.
- Stair #4 involves a variety of paid advertisement strategies with product changes and the introduction of new onboarding features. We're in the middle of this stair and hope it's multiple years before things breakdown again.
Don't give up
I know it sounds really cliché, but the #1 indicator of success is doing the really boring stuff day in and day out and making incremental improvements. As the weeks, months, and years pass by, you will slowly gain domain expertise and start to see the gaps in the market that can set you apart from your competition. It's so hard for founders to stay focused and not get distracted so I would say it's equally as important to have co-founders who hold each other accountable on what your collective goals are.
How GPT is changing everything
I could write pages and pages about how GPT is going to change how the world operates, but I'll keep it specific to our business and chatbots. In 2021, we built an industry specific AI model that did a great job of classifying intents which allowed us to train future actions during a chat. It was a great advancement in our customer's industry at the time. With GPT integrated into our system, that training process that would take an employee hours to do, can be done in 5 minutes. The model is also cheaper than our own and more accurate.
Because of these training improvements, we have been able to conduct research that is allowing us to leverage GPT models like no one else in the industry. This is both in the realm of chat and also training during onboarding. I really want to refrain from sharing our company, but if you are interested in seeing a model trained for your specific company or website, just PM me your link and I'll send you a free testing link with a model fully trained for your site to play around with.
Where we are headed and the dangers of AI
The level of advancement in AI is not terribly dangerous in its current state. I'm sure you've heard it before, but those who leverage the technology today will be the ones who get ahead. In the coming years, AI will inevitably replace a large percentage of human labor. This will be great for overall value creation and productivity for the world, but the argument that humans have always adapted and new jobs will be created is sadly not going to be as relevant in this case. As the possibility of AGI becomes a reality in the coming years or decades, productivity through AI will be off the charts. There is a major risk that human innovation and creative thinking will be completely stalled...human potential as we know it will be capped off and there will need to be major economic reform for displaced workers. This may not happen in the next 5 or 10 years, but you would be naïve not to believe the world we live in today will not be completely different in 20 to 30 years.
Using AI to create deepfakes, fake voice agents, scam the unsuspecting, or exploit technical vulnerabilities are just a few other examples I could write about, but don't want to go into to much detail for obvious reasons.
Concluding
If you found the post interesting or you have any questions, please don't hesitate to ask. I'll do my best to answer whatever questions come from this!
*EDIT: Wasn't expecting this sort of response. I posted this right before I went to sleep so I'll get to responding soon.
Vibe Score

0
Sentiment

1
Rate this Resource
Join the VibeBuilders.ai Newsletter
The newsletter helps digital entrepreneurs how to harness AI to build your own assets for your funnel & ecosystem without bloating your subscription costs.
Start the free 5-day AI Captain's Command Line Bootcamp when you sign up: