Have you tried AI?

Compliance shouldn’t be a full-time job. With Vanta, it isn’t.

Vanta is perfect for startups low on time and resources. Using AI and automation, they get you compliant FAST so you can close those hefty enterprise deals that VC’s drool over.

Plus, Vanta scales with you at every stage. That’s why top companies from startups like Cursor to enterprises like Snowflake choose Vanta.

Tune into Vanta’s product demo on November 12th to see it in action.

Speed Up Your Growth

As anyone who has spent more than 5 minutes on X would know, VC’s are entirely AI pilled right now.

A few months ago, we met with some of our investors to chat about the business. The entire conversation had a pretty consistent theme: what can you replace or improve today using AI?

On the surface I believe that’s a totally fair and helpful discussion to have. Internally at beehiiv we do use dozens of different AI tools and solutions to scale more quickly and do things that were otherwise time consuming or impossible.

And in theory, these VCs have a bird’s-eye view of it all. They are constantly being pitched the latest AI tools and capabilities, and also oversee a portfolio of companies who have likely implemented some of these solutions to much success.

But in reality, there’s a massive disconnect. Most VCs have never actually built a business of their own—let alone sat in the operator’s seat.

The game on the field looks a whole lot different than these happy path demos these VCs are pitched. By the way, “happy path” is engineering speak for an ideal user flow where everything works as intended (this rarely happens in the real world).

And look, I get it. From the VC perspective, I imagine them hearing about some hype new AI security startup with “legit the most cracked founder ever.” Apparently a16z, Sequoia, and Benchmark are all interested in leading their round.

When they finally get to meet the team, they watch a perfectly crafted demo showcasing how this new tool can do the job of an entire security team. That, coupled with a hockey stick revenue chart gives the impression that everyone is adopting this new tool successfully.

This VC saw the future and wants to help their portfolio companies. Is there anything more helpful than suggesting something that would in theory save tons of money (i.e. reduce headcount) and in this example, improve security performance (i.e. efficiency)?

Their heart is in the right place, but the truth is, the capabilities of AI are just not there yet for most of these use cases. Nor have they ever implemented the solution themselves, thus can’t speak from experience. It’s all hearsay.

All they know is that the happy path demo worked when presented by the founder, revenue is seemingly growing, and all of the top funds are interested in investing.

Back to our meeting a few months back, the topic du jour was about support. We have an incredibly talented support team, but ever since we really began to scale the company it’s been a constant battle to provide timely support to all 50,000+ of our active users.

We aren’t alone in this, it’s a growing pain most startups experience.

These investors insisted that we were living in the stone ages, and that AI could entirely relieve us of our duties. The hope was that we could guarantee near-instant responses to our customers and free up some bandwidth on the support side to focus on other initiatives.

We took their recommendation and spoke to a dozen of the hottest, most highly recommended AI support tools on the market. We selected 4 finalists who each built an entire POC for us, fine tuned the model based on our internal data and support docs, and pretty much guaranteed success.

The results were total shit. None of them could handle a fraction of the complexity of inquiries from our users, nor the simplest tickets.

After this exercise, I actually think that entire industry is a house of cards. Perhaps those tools can handle a simple return request for a D2C startup, but there’s a long way to go before they can troubleshoot our users’ automation journeys or bespoke integrations with third parties.

We were unfortunately impacted by the AWS outage two weeks ago, and had a revolving door of downtime between some of our different services. The impact on users was changing by the minute, and required a ton of insight from our engineers to guide the comms for our support staff.

I would love for an AI agent to assess the performance of our vendors, understand the intricacies of our codebase, analyze our dependencies, follow all of the internal communication happening in Slack, empathize with the bespoke actions our users are trying to take, and formulate an accurate response as to how to best assist them.

The AI just isn’t there yet. We wasted a total of two months going down that rabbit hole only to learn first hand that those solutions fall well short of expectations relative to how they are positioned.

The game on the field is different.

Right now we’re hiring a VP of Growth, GTM Engineer, Vibe Marketer, and Partner Lead. Translation: we are about to turn on the growth jets and really accelerate.

One of our investors was confused as to why we would be hiring for these roles, referring to these job functions as “going extinct.” He instead sent over the following chart and suggested we just use these instead.

Oh fuck me sideways, I didn’t know we could just sign up for one of these AI tools and suddenly have thousands of new paying customers. I’m such an idiot for not doing this sooner!

We’re currently rebuilding our onboarding funnel, testing variations of our free trial, optimizing our in-app upsell modals, building lookalike clusters of our highest value users, scaling new creator partnerships, testing new creative and acquisition channels, and launching new product-led growth initiatives.

Or at least, that was the plan before learning that AI could do all of that and better for just a few hundred dollars per month. Instead I’ll be double fisting piña coladas in Hawaii, occasionally checking in on my army of venture-backed AI tools.

I’d argue that even the solutions that work well become far less impactful as their adoption proliferates (for growth particularly).

When companies discovered they could use AI to generate thousands of articles quickly, they realized it was an easy way to boost their SEO rankings and attract significant website traffic. That was until everyone did the same thing and Google updated their algorithm to penalize them.

Perhaps sending thousands of AI-generated outbound emails worked well for the earliest adopters, until every company adopted the strategy making cold outbound more ineffective than ever. AI-generated video ads enabled more startups to compete on Meta, driving up CPMs and making the platform more expensive for everyone.

Now the purpose of this post wasn’t to bash our investors (great people). I just wanted to keep it real for other founders who may be wrestling with the same reality.

The entire venture community has made a massive bet that AI will work as promised. It’s that promise that allows them to raise larger and larger funds, paying them higher fees, and keeping this party going. There’s incentives all the way down for them to encourage this adoption and keep the narrative alive.

And look, I’m not a boomer. Like I said from the jump, we use dozens of AI tools internally and I’m incredibly optimistic about the trajectory of the industry.

There’s just a massive disconnect (today) between the AI solutions investors are shilling from the sidelines, and what is needed to actually operate in the arena. Build accordingly.

If you enjoyed this post or know someone who may find it useful, please share it with them and encourage them to subscribe: mail.bigdeskenergy.com/p/have-you-tried-ai