Back to Research

AI in SaaS

7 months ago 6 mins read
Adam Schoenfeld
Adam Schoenfeld

*I'm proud of* *these G2 reviews**.

Hoping to see 1 more before Thanksgiving.* 🦃



AI features have very quickly become table stakes in SaaS.

We all see it happening.

Today I analyzed 17K SaaS companies to better understand how & where it's happening.

Overall, a whopping 33% of SaaS companies now promote *some* AI features or capabilities.

📊 **This graphic shows a quick breakdown** (keep scrolling for details)

AI isn’t new.

But the Generative AI gold rush kicked into full gear following ChatGPT’s launch (11/30/22) and the growing availability of LLMs throughout 2023.

What is the state of AI in SaaS today?

Here's what stood out in our data.

Front-office SaaS loves AI

Let's start by looking at the occurrence of AI features and capabilities in SaaS companies by category.

To do this we scan the company homepage and LinkedIn description to look for AI specific product messaging, then we break it down with Keyplay secondary categories.

Customer Support tops the list -- 53% of CS tech companies are promoting AI features.

Sales tech and mar-tech players are right behind at 45% and 41%.

Taken together, these categories make up the front-office use cases.

It makes sense that Generative AI is so compelling in this part of the enterprise.

LLMs are unlock all kinds of new efficiency by generating content, ads, emails.

support answers, and personalized experiences.

It’s a compelling story for sales, marketing, and CS.

The bend toward front-office is little different than the shift from on-prem to cloud.

Much of the first wave in cloud was back-office – ERP, CRM, HR platforms, payroll, finance & accounting.

Think Salesforce, Concur, Workday, etc.

Gen AI has infiltrated every category all at once, but seems to be coming on stronger in the customer-facing functions.

Incumbents are NOT sleeping on AI.

Here's how AI messaging looks by founded date.

We can use this as a proxy to understand incumbent and challenger dynamics in the AI race.

When the first wave cloud players emerged, on-prem incumbents were stuck.

AI has a different dynamic.

While the data shows startups adding AI at a higher rate, the incumbents are very much in this game.

The gap is smaller than you might think.

Even looking at the SaaS companies founded in the late 90s, we see 25% promoting AI features and capabilities.

For startups and challengers, the obvious use cases and GPT wrappers won’t create an edge.

As Dharmesh said if you just have a “really fancy prompt, that’s unlikely to be enough.”

Let's look at a couple players in the Data/Analytics/BI category as an example.

On one side we have Equals, a hot startup funded by a16z and Craft, reinventing the spreadsheet.

They have a strong Generative AI story with an "assistant" functionality that can help you explore data using conversational prompts.

AI is not the entire product, but it's a focal point of their message.

Next let's go to the far other side of the spectrum.

MicroStrategy is a $7B market cap public company founded in 1989.

They are a legacy software company that somewhat transitioned to cloud.

Like Equals, AI is a focal point of their product marketing.

Also like Equals, they promote an assistant-like functionality that can help you explore data using conversational prompts.

I doubt these companies ever compete directly, but the comparison highlights how some of the first order use cases are likely to be implemented at companies big and small, old and new.

In other words, startups do not seem to have any structural advantages when it comes to integrating AI.

The Customer Support category (where we see AI integration the most so far), also follows this pattern.

The emerging players like Lang, Thena, and Zowie are doing exciting work with AI.

Meanwhile the established players like Zendesk, HubSpot, and Intercom have also moved fast.

In the middle you have rising stars like Ada and Replicant with a AI DNA also continuing to innovate.

Again, these companies are competing for different market segments, but the ubiquity of these first order AI features makes it feel more like table stakes than differentiator.

Investors love an AI story.

Let's look at AI messaging based on how recently a company has raised funding.

I can’t say if it’s causation or correlation, but the more recently you raised money, the more likely it is that your website is talking about AI.

We already know that investors are excited about the AI platforms and enablers like OpenAI, HuggingFace, Cohere, and Inflection.

Here are just a few of the eye popping funding numbers:

Their appetite seems to extend to the application layer, writing checks into AI-integrated SaaS companies across functional, horizontal, and vertical categories.

I’m excited to watch this next generation of AI-integrated SaaS.

The ones that can go beyond a “fancy prompt” have the opportunity to reshape existing markets and/or create entirely new categories that weren’t possible before.

So What?

Understanding AI development in SaaS helps us all build our own product and GTM strategy for this new paradigm.

While everyone is sprinting to keep up, I find it helpful to zoom out on the market dynamics and zoom in on specific examples in adjacent spaces.

At Keyplay we've been fast to integrate LLM capabilities, but we have been slow to make it part of our story.

Currently we use AI to generate derivative data and signals, power our industry classification framework, and as a key ingredient for our Lookalikes models.

We believe AI will be a key ingredient to our success, but also recognize that it's not a silver bullet.

Each day we learn a little more about the opportunities and challenge, betting that a real breakthrough will come in time.

*How are you building, positioning, and selling around AI (or not)?*

Reply or join the conversation on LinkedIn


If you are looking for this kind of data on your customer list, Keyplay can make it happen for you!

As part of our free POC process we can crunch the numbers on your customers list, find lookalikes, and hone your ICP model.

Just reply here with a little background and I'll get the process started for you.




*Have a B2B tech company in mind you’d like us to track?*

*Fill out* *this form**.*