SaaS Conversion: Which metrics matter?

I was in a VC diligence meeting yesterday and one of the partners asked me which metrics we were most focused on.  The answer to that question varies depending on the company and what stage they are at in their commercialization.

The SaaS Conversion Funnel

At a high level, the key metrics to focus on are:

Acquisition: Conversion rate from new unique visitors to new members, sign ups, etc?

Activation: Conversion rate from new sign up to active user.

Retention / Churn: % of users that remain active (or conversely abandon or churn out) over time.

Revenue: Conversion from free active users to paying.

Referral: % of users that come from existing users.

This is of course the Metrics for Pirates framework. I’m grossly oversimpliying things. You can and should go way deeper, but this is all I need for this post.

Which Metrics matter?

While you can and should be measuring many things, your focus should always be on one or two aspects of the conversion funnel. The aspects you choose depend on what stage your company is at.

Early on, your focus is on acquisition. You just need users. Testing your home and landing pages and optimizing for sign-up effectiveness is most important. Do people get it?  Does your web site do a good job of getting them to sign up?

It’s OK if they don’t stay long. You should expect activation and especially retention to be low. Just focus on acquisition. In parallel, start talking to users, both the really good ones and the ones that abandon you. Find out what they’re thinking.

Once you feel you are effective at acquisition, its time to focus on activation. In truth, you will probably tackle this at the same time as acquisition. In my experience, especially with free apps and services, if you don’t catch people and onboard them right up front, you lose them – forever. If they have a problem on setup, they will just assume your app does not work, and leave. So you need to reach out to users on sign up and get them up and running. Focus on getting more and more of your sign ups active and engaged.

In the long run, the most important metric to a SaaS business is churn rate – the rate at which users cancel their subscriptions. The lower this rate, the longer they will pay you money. Thus, the more valuable they are to you. This gives you more money to invest in acquiring them, so you can have more users all paying you money for longer periods of time. It becomes a virtuous circle.

Don’t worry too much about churn in the early stages. Focus first on your acquisition and activation engines. Even once you begin paying to acquire users, expect your lifetime customer value (lifetime revenue less cost of servicing that revenue less cost to acquire the user) to be negative. Just make sure that the deficit is continually getting lower.  Once the lifetime value is positive, you’re ready to hit the gas pedal and invest in growth.

  • Pingback: SaaS Metrics: Tracking and Improving the Right Metrics | OpenView Labs()

  • http://tracker-go.com Michael De Marco

    Mark, what is a reasonable expectation for a conversion rate of a saas product. we have a crm product. please visit tracker-go.com and if you can let me know what an expected conversion rate should be

    • http://startupcfo.ca Mark MacLeod

      Conversion rates vary widely. Overall, you should target at least 5% conversion to paid. But this assumes a highly growing user base and thus is artificially low since most new users are free. If you look at your cohorts you should see higher rates (up to 10).

      Your landing page needs some work. Not totally clear what you do. I'd play with that. Maybe try our portfolio co unbounce.com to help.

  • http://twitter.com/markzohar @markzohar

    Good post Mark. I would add another metric which is important for both SaaS companies as well as consumer Web services: virality or viral co-efficiency. This metric tracks the viral or organic growth of your product. Specifically, are your activated users bringing in more users through viral channels made available through the product itself or via third party channels (e.g., share buttons, invites, Twitter and Facebook feeds/posts, etc.). A positive viral co-efficient is a strong indicator of: a) positive user experience; b) strong product-market fit; c) high utility and/or entertainment value and d) low cost of acquisition. A company with a positive viral co-efficient can invest earlier in marketing and other growth programs as every dollar invested in customer acquisition will yield a multiple of growth. Optimizing your product for virality and tracking viral co-efficiency as a key metric is incredibly important for start-ups.

    • http://intensedebate.com/people/mmacleod Mark MacLeod

      Thanks Mark. And for sure, your viral cooefficient is very important. I do think baking the direct user experience comes before tuning for virality. But I should have talked about this as well. Hugely important.

  • http://intensedebate.com/people/Healy Healy Jones

    Mark, good high level description of important metrics for SaaS companies.

    A couple of thoughts: I think it’s fine for B2B focused SaaS companies to start charging right away. I know you’ve seen a lot of SaaS companies; what is the best way for a company to go from the “it’s free for everyone” to “we now have a paid version” effectively?

    I like revenue churn in addition to subscriber churn. However, it’s a huge pain to calculate, so it’s more of an aspiration thing for me. But the real question I have is: when do you actually count someone as churned? I’ve seen it different ways at different SaaS companies. The problem is you never know when a customer is going to come back from the dead… I tend to be very harsh in counting someone as churned (i.e. they didn’t pay that month so they churned) but our customer service team is able to save a large number of them so I end up unchurning them… Actually, not just revenue churn but subscriber churn is a huge pain to calculate. How aggressively do you try to make this “accurate” vs. just using it as a directional number?

    Finally, when I was a VC, we did look closely at SaaS churn rates because it impacted the LTV (as you mentioned.) I think it is a pretty important metric for startups to think about early on in their life. Have you found that VCs are relaxed when looking at churn rates?

    Sorry for the disorganized comment to your well put together post.

    • http://intensedebate.com/people/mmacleod Mark MacLeod

      RE: Going from free to paid – to me the only way is to keep the free base and add something on top that's paid. If you do a bait and switch and charge for what used to be free that will piss people off. If something is free for a limited time, that has to be clear up front. That is a different strategy altogether. It has its pros and cons. It takes a hugely viral (and simple) product and / or lots of capital to make freemium work. absent one or both of these, you should charge as early as possible.

      RE: Churn, in the telco World, a subscriber is considered to be churned out after 90 days. I do measure 30 day churn, but I don't consider them fully churned till after 90 days of not seeing them around.

      Finally, I can why VCs would place so much importance on revenue churn. It all depends on the stage you invest. If you're a seed investor, you are focused on engagement. If, like Atlas, you series A and up, you want monetization validation, and you're valuing based on per user economics, etc.