SaaS Due Diligence: Using trial data as a leading indicator

Victoria K Peng
9 min readFeb 18, 2016

--

Startups at a very early stage (<12 months) often have a limited amount of data to use to quantitatively assess their product-market fit. With this limited data, what other metrics can founders and investors use to gauge product-market fit and traction to benefit operational efficiency and investment potential? Focusing specifically on enterprise SaaS businesses, one potential option is to use “trial” data as a leading indicator to go one level deeper to understand the potential growth trajectory and where the business is going.

44% of enterprise SaaS companies use trials to get their customers in the door. Paired with the table-stakes metrics of strong MRR (monthly recurring revenue) growth, low churn, a short CAC payback period), trial data metrics can provide another leading indicator for early investments as a “quantitative lens” to anecdotal rave reviews from customers.

Here, we dig into two pieces of trial data can be used as a leading indicator for evaluating early-stage SaaS startups:

  1. Using unconverted trials to predict growth trajectory
  2. Looking at the % of customers who convert before the trial end date to assess spikes in product-market fit

Why trials?

Trials are a particularly interesting as customers who have signed up for a trial have demonstrated strong initial interest, and jumped through all the hurdles to signup. Unlike freemium models, most trials require customers to enter a credit card, a proxy that demonstrates increased interest. In the long-run, by channel, the split of self-serve revenue from trials versus direct purchase can be as high as up to 50/50 making it a critical indicator versus other on-boarding funnel metrics.

If you have an early stage enterprise SaaS company with strong net new MRR growth, low churn, and MoM high degree of trial starts (where the unconverted trials would not be reflected in net new MRR), it could mean the company has a high growth trajectory down the road and is more interesting in the long-run.

1) Using unconverted trials to predict growth trajectory

For startups at a very early stage, the revenue data may not quite be there yet as you only have a few months-sized cohorts. But, if you’re getting a lot of trial starts, even if they haven’t converted yet, provides an interesting positive leading indicator of what may come and points to latent growth opportunity as well as the degree of product-market fit.

To see this, we take a fictional SaaS startup that uses trials, take the new, expansion, and contraction revenue, and then stack unconverted trials MRR on top on a cohort basis (Unconverted trial MRR calculated by # of licenses trial signed up for* ASP) to provide another layer of quantitative depth.

Excluding existing contracts for simplicity sake. Unconverted trial MRR = # of net new trials in the same cohort period that did not convert (converted trials captured in net new MRR)

Stacking unconverted trial MRR on top gives us more information on the growth trajectory of the business by seeing what money has been left on the table through the pipeline of trial starts. This isn’t captured in net new MRR because these trials didn’t convert, but is also not captured in new user growth data because for SaaS companies that aren’t in a freemium model, you are only a user after you have become a customer.

Now let’s take two fictional example companies. Company A and B have similar new monthly recurring revenue, and cancelled MRR month over month and above average trial conversion rates. But, once you stack on unconverted trial monthly recurring revenue, company B has 3x more unconverted trial MRR left on the table (axes are the same here). With this data, you can see that company B has a lot of latent opportunity of unconverted trials in the pipeline that isn’t reflected in its current new MRR.

Given comparable above average conversion rates, Company B’s high number of raw trial starts indicates more market interest and customer awareness and potentially better initial product-market fit than company A even with <15 month limited data. As company B improves its trial conversion rate over time which companies do as they mature, it could have a more interesting growth trajectory in the long-run.

For more granularity, zooming in, we can further split new MRR by channel into “New MRR from direct purchases” vs. “New MRR from trials” (and to keep the new MRR direct purchase vs. trial ratio in mind) before we stack the unconverted trial MRR data . These sub-components help answer where new MRR is coming from and how growth is being driven as well as the latent opportunity in unconverted trial MRR.

Granular breakdown of New MRR from Trials vs. Direct Purchase

From an operating perspective, a startup can leverage this granularity to better improve their product-market fit by understanding why the product got a customer’s attention, but was unable to convert them (e.g. communicating product value clearly, common customer use cases, etc.) as well as develop LTV on a cohort basis of customers that come in via trials versus direct purchases and leverage that data to understand the value of the product to different customer archetypes.

2) % of customers who convert before the trial end date to assess spikes in product-market fit

Another metric within trial data to leverage as a leading indicator is the % of customers who convert before the trial end date to assess spikes in product-market fit.

Say both Company A and B have an overall trial-to-purchase conversion rate of ~30%. Typically, customers on a trial who convert do so on the day the trial expires as an automatic charge (14-days, 30-day period, etc.), but if you dive in deeper to understand the day-by-day layer of conversion, the data tells you something different. Here, looking at a specific cohort of trial customers and which day they converted through the 14-day trial period:

Example conversion patterns — e.g. 4% of teams who signed up for a trial convert on day 0 of the trial.

In this example, Company A follows a typical conversion pattern where most trial customers convert on day 14 after a trial ends. Company B, however, spikes and has 2x the conversion into a paid customer on the Day 0 of the trial versus Company A. This is a signal that Company B’s product is able to get to that “wow” moment much earlier and potentially indicates a higher product-market fit.

Using the % of customers who convert the before trial end date (in this case 12.5%) is a quantitative way to look at rave customer reviews to confirm customer validation. In this case, company B could be a really compelling investment (especially if the ASP of the product is high) compared to company A given the unique spikes in day 0 conversion throughout the trial period.

From an operational aspect, companies can use this framework to gauge the degree of their product-market fit as well as how well they are communicating their product value to customers. For example, one method is to breakdown the exact engagement activity during Day 0 necessary to make the customer successful (e.g. setting up account, provisioning licenses, etc. a la Twitter follow 5 friends magic number) in order to improve the on-boarding flow.

**Note that day of trial conversion is a slightly weaker leading indicator as most customers will probably not convert early during their trial period (<10% of a cohort). However, for companies that do have this earlier conversion, it provides a very strong signal for product-market fit as only a particularly compelling product could drive spikes in this.

Some things to keep in mind:

Trial-to-Paid Conversion Rate — This framework pays less attention to the trial-to-paid conversion rate and more on actual trial data due to the many confounding factors that can dramatically affect the conversion ranging from trial duration to credit card information upfront to the number of licenses selected at trial start to the strength of the on-boarding flow.

That said, based on industry benchmarks, if the overall trial-to-paid conversion rate is 50%+, it’s extremely good, and if it’s below <10%, there may be a broader product-market fit problem. In which case, if the company does not have a strong SaaS metrics to start with, and net new MRR is driven heavily by trials as opposed to direct purchases, would recommend not using the above trial metrics as a leading indicator.

Freemium model — The other common method for SaaS businesses to get customers in the door is the “freemium” model (e.g. Slack, Dropbox Pro, etc.), and in many cases companies use both freemium and trials. For companies who use both freemium and trials (e.g. Intercom, etc.), trial data is another datapoint to serve as a leading indicator for “business” products versus individual premium users (e.g. Intercom Acquire vs. gated free individual), and doesn’t address the freemium portion. In this case, given the unconstrained “time” nature of freemium (as opposed to a constrained 14-day trial duration), and more available user data, MAUs and user engagement data are much stronger leading indicators to evaluate the growth trajectory of the business.

Direct sales model — This framework only applies to self-serve sales SaaS businesses as opposed to a more hands-on direct sales B2B models. For direct sales enterprise models, Jason Lemkin’s Lead Velocity Rate (LVR) metric applies a similar logic with workable leads at a more mature stage. Other similar leading indicators are Marketing Qualified Leads and whether they are increasing over time in order to gauge future new MRR.

Consumer trial leading indicators

This “trials” framework could also be applied for some consumer companies, but as a leading indicator is much murkier given the broad swath of ways a consumer company can engage users. MAUs/DAUs/archetype engagement metrics are much better data-points to draw from.

However, one case is where certain tactics can be used as a proxy. For example, Sprig offers $10 off the first order, effectively subsidizing it — so the first order can be considered a “trial”. Using a cohort analysis, we can split out all first time users via promos as a “trial” within MAU growth accounting. From there seeing the conversion or drop-off rate after the first order or “trial” can indicate —

1 — There is a strong market awareness and interest in the product at the top of the funnel and there is enough stickiness to go through the signup process.

2 — But, the experience is or is not compelling enough whether it is the food or delivery experience. This number can also be captured in the repeat order rate as well.

Like enterprise SaaS companies, we can use this additional layer of detail to split out first order vs. recurring orders to assess the growth trajectory of the business. That said, this is a much weaker leading indicator for consumer companies compared to other engagement metrics.

Next steps

Further validate the above two metrics, using a robust amount of of early enterprise SaaS company data, the next step is to establish the target benchmarks and standardize across companies to use for early stage investment diligence the:

  • Ratio of unconverted trial MRR to new MRR that applies across the board and points to a great growth trajectory (e.g. the equivalent to SaaS quick ratio of >4)
  • Exact X% conversion before the trial end date for successful companies by day X on a cohort basis as well as for each “stage” of company (<12 months, 12–18 months, etc.)
  • Validate that the above framework applies consistently across different trial models (14 day trials, 30 day trials, no credit card, credit card, etc.)
  • Dig into “net new MRR direct purchase” vs. “net new MRR trial” ratio to see if there are any interesting insights there that can be leading indicators
  • Assess potential to build out similar play-books for enterprise SaaS companies using freemium (could be particularly interesting for startups based on open-source such as Confluent, Sentry, etc. to look at the lag in open-source user adoption to paid user on a cohort basis) and the direct sales model

Note: The idea here is to provide another leading “datapoint” to assess early enterprise SaaS companies who use trials. At the early stage, a strong team is critical as well as a market in which there is conviction, vision, core product, and strong numbers in MAU and MRR (>4 SaaS quick ratio), etc.

--

--