Navigating trade offs in your LLM tech stack AI leaders from Google, Cohere, & AI Fund share insights... Register Now x
Article | Read in 4 minute

How to win at enterprise sales for AI startups

Sales experts currently leading and advising AI startups share what it takes to drive enterprise sales and how it differs for AI companies compared to other SaaS companies.

Panelists:

Israel Niezen
CEO & Co-founder of Factored, an AI Fund portfolio company. Previously, he was Chief Revenue Officer at Telescope and co-founded ScopeBrasil.

James Milin
CEO & Co-founder of Workhelix. Previously, he held sales roles at AWS and Google, and was co-founder and chief revenue officer at an AI startup.

R. Paul Singh
CEO of Startup Strategies & COO of The Modern Data Company. A seasoned C-suite executive, Singh has held CEO, COO, CRO, and CMO roles and founded several startups.

Rocky Paap
Chief Revenue Officer at Filtered, an AI Fund portfolio company. Previously, he’s held various CRO and Head of Enterprise Sales roles at firms such as WorkRamp, Envoy, Box, and others.

Moderated by: Fahad Aziz
Founder in Residence at AI Fund, building a sales automation platform using Generative AI. Previously, he co-founded healthcare firm Caremerge (acquired by VoiceFriend).

 


Here are highlights about what the panel of experts had to say about selling AI products:

Fahad Aziz: How is selling an AI platform different from selling a non-AI platform?

 

Rocky Paap: One thing that’s different about selling an AI product is that it’s more interesting to buyers. They’re more likely to ask you deeper questions, like, who builds your models? Are you building your own in-house, or are you leveraging API’s? And if you’re using API’s, which third party applications are you using?

 

R. Paul Singh: The good news about selling AI is that opening the door is easier. Every time there is a new technology that’s sparking a lot of curiosity, there are lots of tire kickers out in the market. The bad news is there are too many tire kickers who are not really buyers. Our job is to figure out very early who is a mere tire kicker versus who is actually going to buy.

 

 

Fahad Aziz:  Does your sales team need to be more informed about how the AI models work, or would you push that to the solution architect?

 

Rocky Paap: My philosophy is that the account executive should know their product very, very well. They should know specific use cases and edge cases, they should know how it stacks up against the competition. But sometimes there is value in job pairing, where you bring the solutions engineer to have a conversation with a technical counterpart. It gives comfort, and it provides some more credibility. The same goes for bringing in executives into sales, particularly on the enterprise side.

 

Fahad Aziz: James, you’re not selling just a platform, you’re also going into an organization and saying, “Hey, this is how to run your business.” That’s a much more difficult sell. How do you go about it?

 

James Milin: I think it’s really important to say, “We have this incredible technology that can increase productivity tenfold in ways that simply haven’t been done before.” It wasn’t that long ago where 75% of work in America was agrarian. When tractors came along, it made things way quicker and made way for new jobs. The same kind of technology revolution is happening with generative AI. It’s an incredible productivity unlock. When we get folks to see that, the sale goes a lot easier.

 

Fahad Aziz: With every organization that sees a need to bring AI into their products, the first question is “build versus buy.” What are the aspects of that conversation?

 

Israel Niezen: Everybody wants to build their own team in house. But guess what? The US doesn’t have enough data scientists and data engineers, so you have to go outside. You have options. You can go to all these other places that are not timezone efficient. But if you want people in your own time zone, Latin America is a solution. That’s just an example of my business. The point is, think about the problem they have, and just educate them on why you’re a good option.

 

Fahad Aziz: ChatGPT and others are threatening white collar jobs. When you are selling an AI platform, does that become a blocker? Does someone in an organization say, “I don’t think that’s a good solution,” because he feels that some of his team members would be let go?

 

Rocky Paap: It’s really, really important to always know who you’re talking to, and what their specific concerns, challenges or goals are. At the executive level, they might be fine with replacing humans, because that’s gonna be a cost savings. But if you’re talking to the human who’s going to be replaced, then you don’t want to say, “Hey, your job is gone.” It’s more like, “This is going to alleviate a lot of the manual tasks that you’re doing on a daily basis. And it’s going to allow you to focus on more strategic initiatives within the company. You can be one of the first to learn how to use this technology, because it’s inevitable. And in 2, 3, 4 years, you’ll be ahead of the curve.”



James Milin: Technology has been advancing for decades, and job growth, the US economy, and the value of human capital is going one direction, which is up. No wise executive walks into a roomful of people with typewriters, puts computers at the desk and says, “Great, we no longer need employees.” He goes, “Wow, I now have this thing that I can be so much more productive with.” What I’m trying to get people to see is you want to deploy this, aggressively and quickly, to really take all the monotonous things you’re doing in your day job and get them done 10 times faster.

 

Fahad Aziz: Can you share examples of sales strategies that really worked?

 

Israel Niezen: Hyperpersonalization. Some sales leaders say, “Send as many emails and make as many calls as possible.” I would rather my sales team and I take as much time as possible to personalize each outreach to a single human. Make your outreach look and feel like it’s fully personalized one-to-one, a conversation with them and no one else.

 


Rocky Paap: Selling through transparency. I want to research the company ahead of time, look at their website, look at their values, look at the contacts I’m talking to, and see what they care about. And make sure that when we start the conversation, we’re focusing on them and what they care about. When we do talk about ourselves, I want to be very transparent. If we’re a small company, I want to say, “Here are the benefits of working with a small company. And here are the downsides. If you like the benefits, then we’re the right company. If you think the downsides are not the right fit, then we’re probably not.” If you lead with that it’s disarming. It’s not like you’re always trying to sell, sell, sell. It’s more like, “Here’s our view of the world. And here’s who we are.” I found that type of transparency builds a lot of trust.

James Milin: Demonstrations. If you can get an AI demo to land and they say, “Wow, you’ve done in 15 seconds what took me weeks of work,” you’re gonna get the sale. There’s no way a sales pitch or reading marketing material would do that.

 

 

 


Q&A

There were several questions posed to the panelists that we ran out of time to answer live, so the presenters all took some time afterwards to answer these questions:

 

What is the best way for us as startups to position ourselves using the potential benefits of AI without quantifiable data of our own?

James Milin: Some incredible research has recently been published on the benefits of AI. I’d lean into these:

 

 

Israel, in the case of hyper-personalization that you mentioned, what is the success rate for this strategy?

Israel Niezen: It depends on many factors, but hyper-personalization, for high-ticket B2B sales at least, is much better than the alternative. The conversion rate for mass non-personalized emails will approach zero, burn your brand, your relationships, and end your domain in spam folders….so better to personalize it and have a few leads convert.

I have two SDRs, they don’t do high volume outreach anymore because I asked them not to, I asked them to personalize. They each get me on about 2-3 calls per week…after a few months of practice and testing of messages, etc. We have a good close rate once I speak to people, so that volume is more than good enough to justify their total OTE (on target earnings) and all-in cost to the company. Each company, with different deal sizes, will have different math to do of course in terms of number of leads, conversion metrics, quotas, etc of course.

 

 

Deloitte shared on their State of AI 2022 report that 29% of companies are finding that AI isn’t providing the value they anticipated when adopting, and this is trending up as hype continues. What are some ways that we can get ahead of this and set realistic expectations?

James Milin: The timing of publication indeed matters (e.g. 2022 vs. now) as it provides context for the information, especially considering the rapid advancement of AI technologies like large language models (LLMs). As AI has evolved to include more general-purpose technologies like LLMs, it has become accessible to a broader range of applications, unlike previous reinforcement learning models which were often costly and highly specific.

 

How can someone without a network get enterprise deals?

Israel: Everyone has a network or can build one. Track down former classmates, friends, friends of friends, chase people at conferences, follow people on LinkedIn, genuinely care about them and their content, engage with them, produce great content of your own, etc. Also, hire people who have networks or make them advisors.

 

 

For those that are new in enterprise sales, what books or resources do you recommend?

James Milin: These are among the greatest books in the field:

  • The Challenger Sale by Brent Adamson & Matthew Dixon
  • Secrets of Question Based Selling by Thomas Freese
  • How to Win Friends and Influence People by Dale Carnegie

 

 

How do you handle the buy vs. build discussion? I.e: “Why should we pay for your product if we can build our own?”

Israel Niezen: Math. Show them the numbers and why building their own makes no sense. Will building their own generate less revenue? Higher cost? Less efficiency? More distractions?

 

As a technical founder, how do I find agencies to help with B2B sales?

Israel Niezen: Hard truth = you don’t get anyone else to do the first sales for you, and certainly not an agency. Not if you are expecting to be the CEO. If you are the CEO, even as a technical co-founder, then part of your job is to sell the first X-amount of dollars yourself (old school people like me say first $1M). If you are not the CEO, then be the CTO and find a CEO who will do the first sales. The process of making the initial sales will teach you a LOT about your technology and your product, how to improve it, what matters, what doesn’t, what features to remove or simplify, how to articulate it, etc etc. You cannot and should not outsource this in the early stages.

WRITTEN BY

Dan Landau