You might have heard the expressions “Data is the new oil” or “Information is the oil of the 21st century, and analytics is the combustion engine.” While I’m always skeptical of such grandiose claims at first, after a few years of getting to work with and around some amazing high-growth tech companies, including our own SaaSWorks clients, I now couldn’t agree more with these statements.
Over the past few years, my co-founder Vipul and I researched the subscription and SaaS industry as we set out to help non-SaaS companies benefit from the SaaS business model movement and vice-versa — help SaaS companies grow more durably like more traditional software and services companies. We wanted to identify and then focus on what specifically separated the name-brand, high-growth tech companies from the rest of the pack.
We interviewed, and in some cases, explored potentially investing in, well over 200 venture-capital and private-equity-backed businesses to best understand the problems in starting, growing and, ultimately, efficiently scaling these businesses.
After the first 50 or so discussions, it became clear that the businesses where data and analytics informed the investment of time and resources had more durable growth and better metrics all around (such as lifetime value, net revenue retention and the quick ratio).
These businesses created a data and analytics flywheel: the more data that was consumed and the more insights that were produced, the greater the thirst for data became, producing better insights and better outcomes. This in turn allowed the businesses’ resources to be invested correctly and efficiently, leading to higher quality and larger amounts of capital, which led to better investments and subsequent data and analytics — wash, rinse, repeat!
You could even argue that it often was not the product or technology that was the differentiating factor in these businesses but the company strategy and execution that was informed by data and analytics. We’ve seen select product-led growth companies use such data and analytics to race past other companies known for their product innovation. Previously in our careers, Vipul experienced this phenomenon while investing in high-growth businesses at Goldman Sachs and Arrowmark Capital, and I witnessed this while growing and scaling HubSpot.
We then questioned the widely accepted statistic that 99% of startups fail (as defined by reaching an IPO or sale of the business in which initially invested the capital plus a reasonable gain is returned to the investors, versus a $0 return). It was clear that data and analytics were core to the success and most tech companies — at all stages — are awash in data.
So why is it that early and mid-stage companies are data-disadvantaged while unicorns and decacorns eat data for breakfast?
To our surprise, the answer was fairly simple: Gathering, aligning, enriching and completing a multi-system data-set is time-consuming, expensive and requires fairly specialized resources, such as data engineers, data scientists (with data visualization skills) and, in many cases, financial analysts. On the surface, this is a $500,000 problem annually.
Most early and mid-stage growth companies have parts of these skill-sets, but 9 out of 10 times (or 99 out of 100, to keep to the startup failure rate), these skills are focused on building products or solutions, not business data and analytics. In fact, many CTOs we talked to clearly stated if they had (or have) such resources, they would allocate these resources to their product team to help understand and build features that increase product adoption and usage.
In the current job economy, if you were to hire dedicated staff for these positions, you are looking at an increased payroll of at least $500,000 annually, and you would have more capabilities than you likely require due to the specialization needed. Let’s break this down expense-wise in cities known for their tech ecosystems: San Francisco, Boston, New York, Portland.
- Data-Engineer: $150,000 annually
- Data-Scientist: $250,000 annually
- Financial Analyst: $100,000 annually
In an early or mid-stage company, most businesses only need 25% of each of these resources which creates an at-odds situation. Most businesses cannot (and should not) rationalize such roles until the revenue and growth rates warrant it. We often see companies as they secure later-stage funding building such teams, which leads to a steep increase (often a step-function) of both a cost/investment and time as these types of resources are difficult to find.
Sharing these specialized roles across departments such as Engineering, Finance or Operations often leads to conflicts in priority, and therefore the investment is never justified. On the flip side, we observed that the companies that become unicorns and larger had C-suite-driven investments in dedicated staffing or re-allocated individuals to produce data, analytics and KPIs that informed the business’s operations. Often this was satisfied by an MBA intern or other multi-skilled resource, which are very hard and often expensive to find.
The disadvantage became clear: you needed to scale in either the rate of revenue growth or the level of revenue to justify and afford the ability to manage data and perform the analytics needed — or you needed a large amount of capital and founders with support from the investors to invest. The companies that become unicorns and decacorns had either or both!
As with most things in the startup ecosystem, this is a solvable problem. The data exists within the virtual four walls, and with commitment — even part-time — the data can be transformed into something meaningful and measurable to start the motion on the flywheel.
Please note, there is a reason that CDPs (Customer Data Platforms) are becoming all the rage nowadays, but we caution companies from perceiving them as silver bullets. They are another platform to manage your existing data where you are largely responsible for the inputs and outputs of those platforms. In our experience, to say “some assembly is required” is an understatement.
CDPs allow practitioners to create a myriad of results (outputs) ranging from health scores to dashboards to reports to lists. All of these “outputs” require care in crafting analytics to then produce insights via reports, dashboards, etc. (visualizations). We encourage a crawl, walk, jog and then run strategy such as
- Identify select KPIs needed to run the business. We’d recommend three to start, but no more than five (see here for five examples)
- Get your data into a centralized location. In all truthfulness, many businesses we observed were successful in using tools such as Google Sheets and Excel before putting the data into a formal database or even a warehouse such as Redshift or Snowflake
- Build the metrics. Even prototyping them first in Excel or Google Sheets and committing to updating these metrics monthly is a sufficient starting point
Once comfortable, include these metrics in your board deck and other forums. They may feel unnerving to start, but this builds the muscle group to create such metrics repeatedly and more easily over time. This also may justify starting to dedicate resources — and even moving someone into this data/analytics function (or hiring from outside your company). At SaaSWorks, we are helping our clients in this area and they're seeing the benefits from our crawl, walk, job, run strategy. Check out our case studies to learn more.
Jim O’Neill is the Co-Founder and CTO of SaaSWorks. Prior to SaaSWorks he was a founding team member of HubSpot and helped the company grow from 5 employees to over 1,500 and from 10 customers to over 25,000. He was named CIO of the Year by the Boston Business Journal in 2015.