Huge thanks to Kevin Salimian and Lone Pine Capital for their data contributions for this series, and to Nick Mehta, Abraham Thomas, Avanish Sahai, and Kevin Haverty for their contributions!
Going multi-product clearly offers a better world for software companies. Adding a second product increases the revenue per customer (duh), which allows you to spend more to acquire that customer, which allows you to grow more quickly and, eventually, increase profitability.
The obviousness of this logic leads to an equally obvious question: why doesn’t every company do it? Despite the clear advantages, relatively few companies are able to pull off the strategy. According to research from Lone Pine Capital, only 25% of the public single-product software companies as of the end of 2016 were able to generate > 20% of revenues from outside of their core offering by the end of 2022.
So again, we ask why. Today, we would offer two deceptively simple answers: 1) Some are not well situated, and 2) It’s far, far harder than it appears.
Let’s start with the latter. The challenge begins with building the damn thing. Multi-product forces you to fragment your development resources, figure out which elements should be common (such as UI or security) and which elements should be distinct (workflow or roles), and who ultimately calls the shots on the roadmap. Voila, you’ve gone from a nimble start-up to the dreaded corporate bureaucracy.
It’s also hard to sell more than one product. You need salespeople to learn about a new product and spend time selling something that is unproven and often lower priced, all while trying to hit their quota. Good luck with that. Sometimes, you’re also asking that sales rep to sell outside of their current relationship set, maybe even out of the same unit. In all things, complexity is the death of velocity, and there are few things more complex than convincing sales staff to risk their paychecks. It’s hard to get the entire company across all departments/seniority levels bought into a theoretical long-term “good guy” (multi product) that involves a lot of near-term pain and work, especially if the single product is doing really well.
Historically, very few companies have grown outside of their core product organically. Those that do achieve it through M&A typically struggle (which requires a longer discussion of why this is inferior for making money that we don’t have time for today).
Going multi-product is hard because it divides your development resources in half, forces you to convince your sales team to potentially initially accept lower paychecks, and splits company focus. So what about the first reason going multi-product is challenging to do? It comes down to where a company is in the stack.
Control Points and The Right to Expand
There are some SaaS companies today making headway in multi-product. It isn’t with some new-fangled go-to-market approach; it’s generally with products that compound on each other, or, as Toast CEO Chris Comparato put it, are “better together.” This is where product two is a superior offering because of integration with product one, and product one is improved by product two. We think this concept is the future, and call it a “platform of compounding greatness,” or PCG.
“Better Together represents the idea that every additional product Toast offers works in harmony with Toast’s core point-of-sale system of record. Each product saves a restaurant time, makes them more money, or enables better decisions because of integrated workflows, a richer data set, and a better customer experience.” - Chris Comparato, Toast Better Together Case Study
Generally, “better together” requires starting in a control point and using four patterns of expansion to sell compounding products within a customer—or in extreme examples, across multiple customers and even to multiple stakeholders.
In our essays on Vertical SaaS, we introduced the concept of a control point. In a small business, these systems have an unfair right to sell everything a customer needs.
There are typically one or two control points, “systems of record.” Usually, one control point is in the front office (e.g., Point of Sale, CRM, e-commerce) – “that drives sales, that grows the business, that serves as the cash register.” And one control point in the back office (e.g., general ledger) – “where everything else reconciles to…” - Full Potential Saas 2020
Control points in horizontal SaaS are similar. Mid-market and enterprise customers will have a more distributed decision-making function in comparison with a single small business owner, which, counterintuitively, increases the number of control points. Each of a corporation's functions or divisions become little software fiefdoms unto themselves and generate their own unique datasets, workflows, or identities—meaning there are multiple control points distributed throughout a larger company.
Where do these control points exist? We look for gravity: account gravity, workflow gravity, and data gravity. This gravity anchors your control point and gives you an unfair right to expand and sell products that compound each other.
“Account gravity – the user/sponsor of the system is the highest-ranking individual in the customer organization; it’s the system that requires the biggest financial outlay, etc. [...] Workflow gravity – the system that all other systems integrate into – it’s where the most users spend the most time. Not all workflows deliver the same value; in my experience, the system of record workflow tends to deliver the most value. [...] Data gravity – the system that creates and holds the most critical information and is the hardest to migrate….” - Full Potential SaaS 2020
Let’s go through each of these types of gravity and review how an operator can utilize them to build a PCG.
Account Gravity
When you are the most important vendor to a customer, you have an unfair right to sell them another product. You’re trusted, you’ve been vetted by procurement, and you have the customer’s attention. These are light compounding benefits, but they are real.
Look around at a golf event and notice the sponsors. There is a reason the finance department approves that spend. Once vendors get customers spending nine figures, you have a lot of budgetary lock-ins and a salesforce that can accommodate every whim (and comp Master’s tickets).
Another way account gravity appears is with integrations. Think of it like this—if you are important enough, any additional product or system needs to play nice with you. For example, the CRM for Tidemark is Affinity, so we are not going to choose an email vendor that doesn’t integrate into Affinity or a data scientist who isn’t familiar with their API.
“In the early innings, you’re going to win some business based on your reputation, assuming it’s a good reputation and your customers love you. Your advocates will help you—but that’s only going to get you so far. In the early days of multiple products, we were winning by having our advocates help us…” - Kevin Haverty, ServiceNow
Account ownership derived from sales relationships is not enough to overcome the product and go-to-market challenges described above. And with modern software moving toward extensibility, integrations aren’t likely to be long-term drivers on their own, either. However, account gravity that is captured by the product can be powerful enough to fuel a PCG.
“The other key point to the compounding is that in some instances (like Cloud), there is a common buying center for all this. So what you're really doing is creating a distribution channel into that buying center, which yields huge efficiencies downstream.” - David McJanett, CEO of HashiCorp
All of a sudden, your product is the distribution channel itself. This is a strong position because you don't have to pay salespeople—you’re in front of the customer constantly, and implementations are super easy. If your product offers distribution, you build an app store; if your product does not, email marketing and aggressive customer success teams upselling will be your best friends.
Account gravity can be used to make a PCG, but honestly, there are other options that are much more likely.
Workflow Gravity: Follow the Workflow, Follow the Money
Workflow gravity is utilized by the most important workflow. It’s where the most users spend the most time and it’s where they get things done. How do you spot workflow gravity? There are three ways:
- The most valuable workflow: The process that generates the most important products or outcomes, delivers the most dollars. The more explicitly the process delivers revenue, the better for the PCG. (More on this in our Follow the Money section below.)
- Starting points: If you own the authoring tool, you have an unfair right to cross-sell downstream functionality. In productivity tools, this is paramount—companies like Canva and Adobe have done this to massive success. You can also look for the workflow where key decisions are made that kick off new workstreams. If your software is the place that decision is made, you can easily own the implementation of that choice.
- User engagement: Finally, the most important workflow is, often, also where users spend the most time.
The canonical case study of workflow gravity is the ERP. Take consumer goods companies, for example. They’ll often use SAP to manage their workflow. Get the temperature or sequence slightly off, and suddenly Diet Coke tastes like Splenda. For those companies, everything needs to integrate into SAP because that is where the mission-critical workflow is. This effect is so strong that there are some large companies still trying to run 100% on SAP despite its sometimes outdated technology.
To further analyze this type of gravity, you can bifurcate workflow strategies in two ways: Follow the Workflow and Follow the Money.
Follow the Workflow
Follow the Workflow (FTW) is centered around a simple truth: software applications exist to automate. As you automate workflows, there are the usual benefits of increases in speed, accuracy, and efficiency—but there is more than that.
Workflows also compound upon themselves. As collaboration increases, more processes are digitized and interlinked. FTW starts by automating the most important processes, then following those to the interlinked processes and people. By connecting processes digitally, you make them better together: you add efficiency by increasing speed and reducing errors/manual labor with digital handoffs. In a perfect world, data from one process can help automate the next.
What’s more, by extending the workflow guided by your software, you can further increase automation (and reduce friction or dual entry) and have an unfair right to sell applications to new functions and users. Even better, your customers will generally lead you to a product and how to sell it to them. Just follow their needs.
“The great thing about workflow is that it’s software that helps people work together. You’re naturally going to start selling into adjacencies” - Kevin Haverty, ServiceNow
So, how do you follow the workflow? Which new processes do you tackle? My interview in 2017 about Atlassian’s land-expand approach has some advice
“...the best option is to look backward at the customers that have actually grown really large and see the path that they took over time. Write it all down and analyze those customers. I guarantee you’ll see trends. If you’re just introducing a new product, build a path but don’t fall in love with it. Customers will go where they want to go, and you need to adapt to that fact.” - Beyond Land, my 2017 interview of Atlassian’s Head of Growth (now CRO) Cameron Deatsch
The second option for workflow gravity is to go where customers really care—their wallets.
Follow the Money
Follow the Money (FT$) is a corollary to FTW, where the workflow is a transaction. These are specialized workflows because dollars are involved, increasing the stakes. Oftentimes transactions require key details, documents, and reconciliation beyond the scope of traditional workflow software. This strategy increases in value when you consider that the fewer systems required for financial workflow, the fewer rekeyings, reconciliations, and delays are required.
As a transaction crosses multiple activities or functions within an organization, a software provider follows the money to build a product for each step in the process. A great example of this is “procure to pay,” where companies like Coupa leverage a procurement control point to sell sourcing, contract management, invoicing, and payment management offerings. Expense management software providers like Concur look to extend from corporate spend to travel spend management. Ramp and Brex in SMB follow the money upward to corporate cards (and get the juicy interchange fees on that spend).
FT$ also makes it natural to extend across multiple stakeholders (in every transaction, there’s a buyer and seller). To avoid rekeying reconciliation, payment delays, and other errors, there is a strong impetus to have buyers and sellers automate through the same set of software and payment rails. The end goal is to have an efficient, straight-through processing across multiple firms. In FTW, it's about increasing the speed of business; in FT$, it is about removing delays. For example, as a merchant with Square, you can pay 100bps to get your money three days sooner.
Another example is Ariba. The company provides procure-to-pay within an enterprise, but added on both supplier management and supplier-pay businesses to sell across both parties (you can read our case study of Ariba here). Bill.com and Melio are similarly building a multi-stakeholder network of payments.
To implement FT$, you’ll want to start with the right customers. You are looking for places where value chains concentrate. A classic example are big payers: insurance markets, banks, the auto OEMs. Because they play such large roles in their respective industry, and because they have the money, they can often set standards—including software usage. Convince GM to use your software, and their hundreds of suppliers will all fall in line.
The final option is to utilize Data gravity.
Data Gravity: Single Source of Truth and Data Enriching Data
Data gravity means having the system that creates and holds the most critical information, and is the hardest to migrate. It tends to form around financial or identity data. Changing general ledger software has been compared to open heart surgery—a customer isn’t going to switch unless they have a heart attack because of their previous provider. Identity works around important stakeholders such as your customers, your employees, and suppliers.
Salesforce’s CRM business is a great example of this. They have all your customer data. You aren’t going to move that unless you absolutely have to (and Salesforce reminds you of that with every renewal pricing discussion). For large companies, a CRM migration can take 1-3 years to execute. It is a big advantage to have the data.
We see a few strategies to go multi-product with data gravity.
Single Source of Truth
If you are the data gravity control point, you’re the system of record, aka the Single Source of Truth (SST), for the most important data. If you are that control point, you have an unfair right to provide additional products that leverage or provide different views of that data.
Financial data
Within a company, financial data is the scoreboard for a company’s performance. Eventually, all plans and projects need to be reconciled to a centralized set of financials. This is a lot harder than it sounds, requiring coordination across divisions, currency, product groups, and functions. As a beleaguered financial analyst would tell you, something as simple as sales (aka revenue, bookings, billings, TCV, ACV, etc.) can confound even the most sophisticated companies. We’ll present a case study on our favorite “Office of the CFO” company soon to describe a SaaS startup that is taking advantage of this opportunity.
Identity
Identity is a special case. The identity of your customers, employees, suppliers is so important that we’re specifically writing a love letter to identity (coming soon!).
Salesforce.com’s first product was an automation that helped salespeople track sales interactions and data with prospective customers. One of its most powerful expansions has been Service Cloud, which tracks service interactions and data with those same customers. Salesforce also acquired ExactTarget and Demandware to track marketing, ecommerce sales interactions, and data to those same customers.
M&A is a bad way of aggregating customer data. Salesforce even paid $900m to acquire Krux, which provided a CDP (customer data platform) to try to stitch it all together. It turned out that merging multiple customer taxonomies, and having more than one customer database, is often worse than having none.
Identity Cross-Enterprise and Multi-Stakeholder
Identity is all about what something is, its key attributes, and the permissions given. This is powerful within an enterprise, but really shines in cross-enterprise or multi-stakeholder examples. A cross-enterprise travel software company would sell to multiple hotels. A multi-stakeholder identity strategy would then sell software to hotel’s suppliers, customers, and employees. The business model may be as simple as data co-op to full workflow automation—it depends on the use case. The key component is that it uses identity data as the link in the chain between all of these different enterprises. To help make it clear, here are some examples of this strategy in the wild.
Other multi-stakeholder options could even be something as standard as a contract. These are an articulation of who owes whom and on what terms. By definition, they are meant to be a single source of truth (although lawyers often confound the issue). They can also be a great launching point for the Follow the Money expansion described above.
Another form of data gravity occurs when data starts enriching other data.
(Embedded) Data Enriching Data
The single source of truth is about having the definitive system of record around critical data. Interestingly, there is a corollary concept: the opportunity to provide third-party data to enrich existing data. When this occurs, it can create network effects as demonstrated in the graphic below:
Let’s go back to the Salesforce example. If your system owns the customer data, you can provide third-party data to enrich your view. That was the thinking around Salesforce buying Jigsaw (a co-op database in contact info) in 2010, but that ended up creating a lot of challenges with customers worrying about the privacy of their customer data.
Another bite at the apple was ZoomInfo providing high quality contact data. It is now adding all kinds of intent and behavioral data to supplement that contact data and improve sales. In marketing automation, identity data merges directly with software by tying leads to the lead generation source. Market automation providers such as Demandbase provide both the site software and the identity data—it even takes it as far as ad serving. Likewise, 6Sense, which was primarily a data player, is increasingly getting into messaging software.
One question we are grappling with is how or if AI changes PCG strategy.
AI
As Patagonia-wearing venture capitalists, how could we go so long without mention of generative AI? We’ll be publishing a longer write up of the impact of AI & Applications, but the question of generative AI is stark with SST and data enriching data.
If you’re pursuing SST, we think you’re generally in good shape. You have data gravity, which by definition is the most important data set, and that data is embedded in your application. LLMs will emerge so that you can extract insights from your data. The future is still being made, but until the power dynamics are more clear, keep your data locked down so people can’t extract value for free.
More generally, AI offers new opportunities for UI advances, personalization, and data aggregation. Just make sure to prioritize AI quickly so a disruptive startup doesn’t raise $B from punch drunk VCs and increase your CAC for a couple of years!
If you’re pursuing a Data Enriching strategy, it may not be as clear if generative AI is friend or foe. You have data gravity embedded with an application, which is a strong starting point, but AI can be a disruptive force. Note the wariness in CBinsight founder Anand Sanwal’s voice:
“The thousands of deep software buyer interviews we have on Yardstiq are also not in any LLM. The data we extract from regulatory filings, that we pay a lot of $ for, is also not going to be available in any LLM” Anand Sanwal, CBinsights March 2023
To be frank, it is still unclear when AI will be a compliment or a disruptive substitute. Our hunch is that in the cases where AI empowers and enriches data aggregation it will end up being net beneficial, but AI may end up hurting you if it subsumes application data and creates tricky data ownership questions.
Bringing it All Together
Each of the strategies discussed today, from FTW to Data Enriching Data, are additive. A software company can build a multi-billion dollar organization using just one, or using multiple attempts simultaneously.
One thing that has always bugged us about investors writing strategy proposals is the lack of actionable insights for founders. So, in addition to the essays we have already published, we will be publishing case studies that break down how growth stage startups used PCG strategy to enormous success. We will use well known names like ServiceNow as well as more under-the-radar companies like OneStream.
The goal for this entire body of work is to help entrepreneurs. Tidemark’s mission is to be incredible partners to world-changing founders—we hope this series accomplishes that for you. If you are interested in chatting through PCG strategy, contact us at knowledge@tidemark.com. To get early access to our case studies, please sign up with the button below.
This essay is the third in a series on becoming a PCG. If you want to come explore with us, make sure to subscribe below.
Are you building a platform of compounding greatness? If so, we’d love to talk. We can share more detailed proprietary research and benchmarks and geek out on the power of this approach to SaaS. You can reach us at knowledge@tidemarkcap.com.
