Building a Compound Company with Parker Conrad

Key learnings from Building a Compound Company with Parker Conrad

This episode’s guest was Parker Conrad, co-founder and CEO of Rippling (prior — founder of Zenefits). This episode of Invest Like the Best was especially interesting because Parker is building Rippling in a way that we don’t come across often when it comes to early-stage companies. Rather than focus narrowly on one product, he is building a suite of interrelated products simultaneously to carry out the operational functions of HR, finance, and IT. He calls this, "building a compound company."

At the time of writing this article Rippling’s most recent funding was a $250M round co-led by Bedrock and Kleiner Perkins, with participation from existing investors Y Combinator, Sequoia Capital, and more. The round valued the company at over $11B. The company headcount has gone from 350 employees in December 2020 to 1,700 in December 2022.

[Learning 1] What is a Compound Startup/Company?

For the past 20 years, market forces have been ripping legacy software vendors apart, in part because of a change in the software delivery model from on-prem to SaaS. This shift was large enough to create a wedge for thousands of software upstarts to try and take their slice of the “cloud pie.” As people started to realize the market opportunity of "SaaS"ifying components of traditional software vendors, VC dollars poured into enterprise SaaS and large portions of the IT budget started shifting to these cloud vendors. During this period, the only thing that mattered for SaaS companies was how quickly can you take a feature from one of these on-prem vendors and create an easy-to-consume/stand-alone cloud service. This created a proliferation of point solutions to the software world (the average enterprise has 100s of unique 3rd party applications). The "legacy" vendors struggled to move their products to cloud-native fast enough and the start-ups of the past 15 years benefited. Parker believes that this shift has created a mindset that the best way to build a company is to focus narrowly on one slice of your prospects' problem and stay laser-focused on that one feature until you've earned the right to become a platform years down the road.

Parker and the team at Rippling have decided to flip back to the old model by instead focusing on building a suite of products right from the start (human resources, finance and IT).

Parker believes that narrow point SaaS has been played out a bit, and there is undiscovered product market fit that lies right beyond the horizon line for companies who can build with the big picture in mind. So Rippling is viewing compound start-ups as something new-come again as they harken back 20-30 years ago when companies like SAP, Oracle, and Microsoft did not have thousands of SaaS vendors nipping at their heels and all have very broad product suites.

Parker's thought is that deeply integrated software and software bundling with a single vendor is going to play out again.

He then goes on to call out four specific ways in which he thinks compound companies are uniquely positioned to win:

1. Your products are deeply integrated in terms of the data that is generated from each tool and how you can stitch it together across product offerings.
2. They are built on top of shared middleware capabilities. You are able to amortize engineering efforts across multiple products. These middleware capabilities are?aspects of your product like reports and role-based access control that are normally seen as “something we need to build,” but not something that is built thoughtfully by point solution vendors. But when you’re building reporting for 5+ unique products, you are more likely to make the proper investment in these middleware capabilities (more on this in the miscellaneous notes section).
3. If you’re a Rippling user with one of their products, you have superpowers that apply to you when you buy your next product from Rippling (you’re already familiar with the product, and the business benefits with quicker ROI). This does not happen when you buy a 3rd party tool that’s new to your employees.
4. You can provide bundled price economies of scale as customers bring in more of your products to their organization. Example: Slack seems to be the more desired product by end users, but Microsoft Teams has great market share because it's being bundled into enterprise-wide rollouts of Azure and O365 (Teams is basically free to that Microsoft customer).

Taking point four a little deeper, Rippling doesn’t go after Enterprise accounts as much as the traditional software vendor (the unit economics don't work well for most vendors in terms of sales and marketing spend to acquire a small customer); but for Rippling, because they have so many product offerings, they often have very high wallet share with a customer even as each product might not bring a very high annual spend. Example: Rippling can go in for $10,000 ARR with one product that would not be profitable for a traditional company because they know in all likelihood they are going to be able to cheaply cross-sell more products as time goes on (they have 24 SKUs, so there are many products to attach with).

[Learning 2] The “Go and See” Leadership Principle:

Out of all the learnings from Invest Like the Best episodes I've written about, the “go and see” leadership principles seem the most intuitive, but if I had to guess, is the least well practiced as companies grow to over a couple of hundred employees in size.?

Parker gives the example of Albert Strasheim (Rippling’s CTO/SVP of Engineering), who is running an 800-person engineering organization, so you’d assume he’s removed from the front lines of writing code, but if you look at GitHub pull requests, Albert is right in the middle of the 800 people org and he’s running the whole team. It would be easy for a CTO to say that writing code is not within his or her job description, but Albert seems to see enough value as a leader to get deep into the weeds of his business.

Another great example comes from Matt Plank (Rippling’s VP of Sales). When something is off in the sales org, the first thing Parker is asking Matt is, “how many Gong recordings have you gone and listened to?” He’s expecting that Matt has already listened to 10 calls within the problem area. The idea here is that anecdotes are so much more powerful than high-level summary data. You could try and make assumptions as to why your proof of concept win rate is down 20% or you could go look at the recent lost POCs (listen to the calls, read the Salesforce notes, and ask the AE about the value streams the prospect was trying to align to). This is go and see.

One final example comes from Parker himself (my personal favorite). He’s the CEO of a 1,700-person company and he’s using the Rippling product every day for Rippling itself (managing HR processes). He put himself as a global admin for certain areas of the business so that he needs to keep using his own product to run their business, and if things are taking too much of his time as the CEO, that’s an opportunity for product and engineering to improve the product for all of their customers.

For Rippling, the CEO/Founder knows the product better than anyone (internal and external).?How often is this the case at the scale of a 1,700-person company? Parker wants to get nitty gritty in the details and he expects other employees to do the same. This is a cultural mindset, and when other leaders see top executives exhibiting the “go and see” mindset, it’s hard for them not to consider if they should be doing the same. It seems non-scalable to do this, but there is more ROI in the anecdotes from the front-line stories than in the dashboard view (which is easier and quicker).

[Learning 3] The Unified Meta Schema (Rippling concept):

The unified meta schema concept is the idea that data natively being includable and joinable across an entire platform, with the employee as the focal point of the data schema for Rippling. Traditionally, data gets spread across different systems, and the more this happens, the more it becomes a data operations challenge to compile and make data usable for the business. Rippling has made a concerted effort to make it really easy to pull and push data around.

An interesting example of this is how Parker talks about workflows, where a system might know who an employee’s boss is, but does not as easily know who the Director of the finance department is and this person needs to be in the approval flow, so this approval becomes a little more cumbersome. Rippling has gone through the trouble of really thinking critically about how all this data joins together from a business process perspective and not just at the database level.

Another example is Salesforce, which is very good at storing data. They are very good about letting you know how leads and prospects map to accounts, how territories are defined, and letting you build out workflow automation. But what they don’t have is giving customers the ability to make data actionable and real-time (in part because of a cumbersome data schema).

To have an impactful/unified metadata schema, you want to focus on giving customers ways to do things with your product that you never intended them to be able to do. This means you are allowing for flexibility for customers and can learn from your customers on new angles to productize for the rest of your customer base.

Parker gives an interesting example of this with Rippling’s spend management product.?They had a company holiday party and wanted to be able to issue cards for per diem for whoever was going to be traveling more than 40 miles to get to the party. Parker was able to send a survey to see who was planning to attend (their survey product), check to see which of the attendees lives more than 40 miles away from the party (their HR product), and then use their corporate card product to issue the per diem. Because each of the products is connected via a unified data schema, Parker was able to get this sub-set in a matter of clicks. If you don’t have a unified meta schema this becomes an exercise of taking the data from each system and then joining them in Excel and uploading them back into the corporate card product (a very manual process).

Miscellaneous Notes:

• You should focus on ways that you build in a specific way that is better for customers, not just what can make you short-term revenue. Parker gives the example of how many times people have come to him and said “why don’t you turn Rippling into a bank.” For him, it has never been clear why they would do this. Becoming a bank would be really good in the sense that banks make a lot of money, but the customer would not get any clear value from this. So he calls out that by doing this, you are picking up cash/revenue, in return for a lack of focus on providing customer value. Instead, Rippling looks at the problems their customers need to solve and picks which one they will have a differentiated product advantage because of the make-up of the current offerings that customers have access to through them. Or what problems they could solve for their existing customer personas, as compared to a product that would require them to break into a new team. Example: they have thousands of HR teams using their solutions, so why would they prioritize a solution that was suited for an operations team if they have other options that would be suited for HR teams? The effort to cross-sell would be much higher to operations. Seems intuitive, but I've seen this done wrong, and the cost of sales is much higher than expected.
• And lastly, back to the middleware layer for Rippling; product components that end up getting built into many different solutions — for example, reporting/analytics. By having a compound start-up you can abstract out these middleware capabilities across your various product offerings so that your reporting/analytics is much stronger than reporting/analytics in the single product-focused company that begrudgingly built reporting because “the customer kept asking for it.” Rippling goes all in on reporting and then amortizes that work across many of its product offerings. They want their analytics to rival those of Tableau and Looker, and almost no other company is going to have analytics this strong. And the extra investment pays off.