Understanding Software as a Service

The following article appeared in the Australian Financial Review on July 03, 2018.

Software as a service (SaaS) companies have been gobbling up business, and with good reason. Software sold as an ongoing service not only removes a meaty upfront cost for customers, it also meets them on their own terms given the proliferation of devices, work environments, integrations and desire for real-time connection.

Likewise, investors have been gobbling up their shares. Again, with good reasons. SaaS businesses tend to have sticky customers, especially on the enterprise side, and many are growing at healthy rates.

Not all SaaS companies are created equally, though, and investors would do well to understand and track their key performance metrics. Unfortunately, there are not universal approaches to these measures, but perhaps the following primer will help.

The first key metric is lifetime value, typically shortened to LTV. The concept is simple enough: how much gross profit does a business expect to generate over the life of the average customer? Naturally, higher is better, but what’s more important than the absolute level is the direction. The calculation is a direct output of gross margins (which are a proxy for pricing power) and average customer lives (which are a proxy for customer loyalty) and it’s a very good sign when they’re improving.

The devil is in the details, though. For example, most companies I come across do not account for the time value of money in their estimates of lifetime value, even though the reality is that future-year profits are worth less in present terms than current year profits. Also, the assumptions a company uses around retention rates can have dramatic impacts to LTV.

Even stickier is that companies report retention rates in an even wider array. Some report purely based on headcount, say the number of customers retained after 12 months. A typical SaaS business focused on small to medium enterprises (SMEs) might retain more than 80 per cent customers year to year while ones focused on large enterprises can climb above 90 per cent retention. Big picture, the more mission-critical the product and the deeper and more costly the integration, the higher the retention rate.

Different approaches
Other SaaS companies, though, base the measure on revenue retained from the year-ago cohort net of customer losses. This approach is closer to the mark in terms of the true economics of the business, especially for SaaS companies which charge customers based on usage. But the differences in approach must be considered when comparing SaaS companies.

Another key metric is customer acquisition cost, typically shortened to CAC. CAC is key because it tells investors how much a company has paid to acquire its customers. Once again, though, it’s hard to compare CACs across businesses because they typically do not reveal the finer points of the inputs.

Also, while CAC will tell investors a great deal about a broad cohort of acquired customers, companies typically don’t share enough data for investors to discern the costs of, say, an organic, inbound acquisition versus one that was acquired through costlier external marketing.

Still, investors can take a few key points to the bank when it comes to CAC. For starters, while it is normal for CACs to rise over time at a scaled business as market share increases, it’s very troubling if CAC is rising faster than LTV. Also, savvy investors will typically want to see LTVs that are a few times the size of CAC, showing the business is earning attractive returns on incremental marketing spend.

These metrics aren’t the only ones that matter with SaaS companies, and there is a lot more to delve into for curious minds. Still, it makes for a good starting point, and I’d encourage investors of all stripes to lean on companies to provide more detail on the underlying assumptions companies make in their calculations.

Joe Magyer is the chief investment officer of Lakehouse Capital. This article contains general investment advice only (under AFSL 400691).