For the 3rd instalment of SaaS Math, we’re going to dig into the 1st step in the conversion funnel: acquiring users. For this and all posts on the conversion funnel, I recommend you read/ watch Dave McClure’s excellent talk on Metrics for Pirates.
As the name suggests, acquisition is all about driving new users, readers or buyers depending on whether you’re running a SaaS, media or e-commerce business respectively. We’re just focusing on SaaS here but many of these principles flow into the other business types.
This is not a ‘how to’ acquire users post. Though I will share some thoughts on that. Instead it’s about the math behind acquisition. What and how to measure when it comes to acquiring users.
The image to the left is a mockup of a custom dashboard we built for a startup I used to work with. All numbers are fictitious. While dashboards are great for getting a quick sense for things, when it comes to measuring your conversion funnel (from web site visitors to users to paying customers and referrers) you want to be looking at data over time vs. a snapshot.
When you are working “in” your business, making changes to copy, A/B testing, updating your application, etc, etc, then you should be looking at your data daily. ABT. Always Be Testing.
But, when you are working “on” your business; i.e. sitting back to assess performance and make decisions weekly and monthly data are better.
What To Measure
There are some pretty basic things you should be measuring when it comes to user acquisition:
Traffic: New unique visitors (UVs) during the period. It’s important to isolate for new vs. returning visitors as only new visitors (or more accurately, visitors who have not already created an account with you) should be considered in determining conversion rates.
Conversion rates: UV to Sign up: What % of new UVs create an account?
Changes: In all aspects of your funnel you need to be very sensitive to changes from prior weeks. If, for example, 10% of new visitors usually create an account but only 8% do this week, then something may be wrong. Either some new web site copy is not working or new acquisition channels are not, or something in the flow to sign up is broken, etc. All changes, positive or negative, need to be fully understood.
Growth: This is related to changes, but you also want to look at week over week growth in web traffic, new users, etc. This will give important trend data for business planning.
Source performance: You want to use a large number of paid and unpaid traffic source. One of our larger portfolio companies tracks 75 separate sources for users. Obviously you need to know which source or channels perform the best so that you can double down on them.
The challenge in measuring source performance is that a visitor may come to your site several times before deciding to sign up. This is especially true if your service is expensive or if there is no free version. So you cannot simply take the source of the last visit before signing up as the one that generated the new user for you. You need to be able to track back to the first source.
Google Analytics alone will not do this for you. My companies have used scripts and logs to properly match up new users to their true source. There may be some off the shelf packages that do this well. If so, I’d love to hear about them.
Targets: The best performing businesses tend to have clear targets. If you don’t set weekly targets then how will you determine if your actual results are good? Always set targets for each aspect of your conversion funnel and make someone in your team responsible for each aspect.
What not to measure
When it comes to acquisition, the one thing to not do is focus on “vanity metrics“. Don’t spend time think about your total web traffic. Only focus on how it converts. Similarly, don’t focus on total user count, but instead focus on new user acquisition and how fast your user base is growing.
How to measure
I’ll be digging deeper into some great services like Kiss Metrics, mixpanel, RJ Metrics, etc in the future. These can all be great tools for SaaS startups. The starting place obviously is Google Analytics (GA). Properly setup with granular goals for each channel, GA can be powerful. But it’s not enough. You will also need some scripting or other tools to figure out true source performance (see above) and will likely want to be running a number of custom database queries each week.
I always recommend having one analytics package for your website and a separate one for your application.
This simple example below puts all this stuff together:
I’m probably leaving some stuff out, but that’s the high level on SaaS acquisition math.