Why building a "compound startup" might be the next, great, non-obvious SaaS play

This post was originally published on March 8, 2022 for the Angular Ventures newsletter. Subscribe here to receive all new Angular blog posts, data reports, and newsletters directly to your inbox.

For the past decade, conventional startup wisdom has been to focus. Find a narrow problem, solve it and you’ll be rewarded. And for the past decade, that’s been pretty good advice. The world is full of excellent point-solution SaaS products, many of which unseated inferior modules in old-school enterprise software suites.

But things have gotten a bit out of control. I was chatting with a good friend of mine the other day who is starting as the VP Sales at a fast-growing startup. The number of products he intends to purchase is mind-boggling. Outreach, Calendly, ChiliPiper, Gong, ZoomInfo, Clearbit. Some sort of lead scoring tool he hasn’t decided on yet.

The monetary cost to this is steep (he’ll be spending at least $350/month per rep, not including Salesforce!). But there’s an organizational cost as well. He’ll need to hire a RevOps person to tie everything together, as well as get time from engineering to ensure every tool has the right data flowing in and out.

If you’re a customer, this is hugely painful. And if you’re an entrepreneur, you’re left competing for budget against countless other tools no matter what angle you take. It’s not 2010 anymore.

With that in mind, I recently happened upon this excellent talk by Rippling founder Parker Conrad on the virtues of building “Compound Startups” (a.k.a. companies with a multi-product portfolio rather than a point solution). In the talk, Conrad explains why the “focused” approach to building a software business may be outdated, and makes a case for the strategic advantages to building a compound startup like Rippling.

I found the talk to be quite thought-provoking, and immediately started to think about how I would go about identifying a good opportunity for a compound startup.

Here’s what I came up with:

First, you need an “over-tooled” business process that you can wholly own. You’re looking for the customer pain my VP Sales friend is facing. A single business process that requires too many tools that are too expensive and cause too much overhead.

Second, you need a foundational data layer. Look at Microsoft, Salesforce and ServiceNow, the three largest enterprise software companies in the world depending on the day. They are each the single source of truth for a different type of critical business data:

• Employee data (Microsoft Active Directory)
• Prospect and customer data (Salesforce)
• Customer and employee requests (ServiceNow)

This data layer is key to their success. They can build their products on top of that foundation, and seamlessly share relevant data across the product portfolio. Their solutions may not be strictly better from a user experience perspective, but that’s not the dimension by which they’re competing. They win by reducing complexity for their customers. Their customers don’t need to think about how to get their customer data, or employee data, or support data, into this new system. It’s already there.

And by offering many different product SKUs, they can take advantage of bundled pricing, meaning they can price differentially and extract value more efficiently. Said another way, they can maximize the price of the bundle, but undercut the price of each SKU, enabling them to compete ferociously and still come out ahead. Just look at how Microsoft Teams has been able to compete with Slack.

That’s not to say this approach is easy. If anything, it’s much harder. But if you can pull it off, there’s a much bigger reward, as well.

After more than a decade, the point-solution SaaS playbook has become so standardized, and the revenues, margins and churn so predictable, that VCs invest as if the risk has been all but eliminated. I can’t help but wonder, then, if this is the moment when everything changes.

So, while I can’t tell you what the next great compound startup idea will be, I’m officially on the lookout for it. If you think you’ve found it, I’d love to discuss it with you!