Isaac Madan
Raising a Series B is never easy, and it’s become exceptionally difficult in the last year as the venture spigot slowed to a trickle. But it’s a different story if you’re an AI startup, right? After all, VCs are throwing money at the AI space.
Not so fast.
Already, the headline-grabbing funding rounds for generative AI companies are beginning to slow, and with almost every startup calling itself an “AI company,” it will become more difficult for true AI startups to stand out.
For founders trying to raise a Series B for their AI startups in the next six to 12 months, a more challenging fundraising environment likely awaits.
So what can AI founders do to raise a Series B when AI is everywhere? As a founder who raised a $40 million Series B from top investors in August last year, I can share a few strategies that worked for us.
Convey the “why” behind AI
You may feel like your startup is all about AI, but is it really? Countless startups are looking for ways to incorporate AI into their products. There’s nothing wrong with adding AI features, but if that’s all you’re doing, then claiming to be an AI startup just for the sake of it will diminish your credibility.
If AI is essential to your solution, you should be prepared to explain the measurable impact it has on your product offering. Do your models generate a few points of improvement over the best available baseline, or does it represent a significant step-function leap from the status quo?
You can convey AI’s impact on your startup in several ways, including quantitative metrics like model performance, business value measures like return on investment (ROI) and total cost of ownership (TCO), and qualitative proof points like case studies and success stories.
But impact alone is not enough. OpenAI’s domination of the AI space poses a threat to countless startups, particularly those that act as wrappers to public models like GPT-4. In today’s hyper-competitive landscape, it’s important to articulate how your startup stands apart from major players.
For instance, what moats does your business have? Do you have a valuable, proprietary dataset? A unique business workflow? Domain expertise? These are all critical ways to communicate your startup’s competitive edge.
Establish ironclad credibility with investors
After being inundated with AI startups, investors have become savvier about identifying what is and what isn’t AI. With the accessibility of large language models (LLMs), building an “AI startup” has become much easier and cheaper.
Collecting and labeling massive amounts of training data is no longer a mandatory prerequisite. Now, following the advent of pre-trained LLMs, these tasks can be done with increasing ease. That means many startups have rolled out AI features to tie into the prevailing market trend and seek a bump in valuation.
The best way to distinguish your startup is by demonstrating your credibility and handling technical diligence early on in the investment process.
You should be prepared to answer the following questions:
- What AI expertise do your team members have across model development and deployment?
- Can you describe your model architecture? If applicable, which pre-trained models are you using, and what techniques are you using to enhance the models?
- What datasets are you using for training and fine-tuning? How are you growing these datasets? How are you approaching customer privacy?
- How do you measure the performance or accuracy of your models? Do you have metrics or benchmarks to share? What’s the performance of the best available baseline?
- How stringent are your accuracy requirements? For example, will your model add value to customers if it’s 80% accurate, or does it need to be 95%+ accurate?
- What tasks can your model perform most accurately? What are your model’s limitations?
This last question is of paramount importance. By knowing where your models shine and where they don’t, you will demonstrate not only technical depth, but also integrity.
Get specific about your GTM formula
In a frothy market, you can raise capital to pad the balance sheet without necessarily having an expansion formula or a clear sense of how you’ll spend the money.
However, in today’s environment, assuming that you’re building an AI application (as opposed to conducting AI research), you’ll need to have compelling business metrics to demonstrate product-market fit along with a plan that charts out the growth path ahead.
First, make sure you know your metrics inside and out. That goes without saying. Data is a great way to “show” versus “tell” why your startup is a great investment. Pore over current SaaS benchmarks and evaluate how you stack up against them.
This will help you determine how competitive and investable your business is, as well as how challenging the fundraising process will be.
Once you have your metrics ready, it’s vital to have a clear plan as to how your investors’ capital will translate to responsible and efficient growth. What levers will you pull to develop new products, increase sales, or forge partnerships? The more specific you’re able to be about your go-to-market (GTM) formula, the more confidence you’ll build among potential investors.
In addition to the above, there are several other unique considerations imposed on AI startups. For instance, full-stack AI startups may be selling “work” in terms of labor displacement, optimization, and automation.
If you are automating work, be prepared to explain the economics in detail. Moreover, building and training AI models can be expensive; significant computational and labor resources are necessary for collecting training data, labeling data, deploying models, and more.
Today’s venture market favors capital efficiency, so ensure you have a deep understanding of the cost structure and resources needed to scale, as well as the impacts that scaling will have on your startup’s gross margins.
It’s no secret that AI is the darling of the VC world right now, but the winds of change are swift, and there are already signs the AI funding boom time is slowing. Founders looking to raise a Series B in the coming months should focus on fundamentals. After all, hard work, not hype, is what generates customer value.
Comment