Mike McDerment from FreshBooks gave  a great presentation on the basics of web application marketing metrics. He focuses on the metrics, systems and reporting that all companies should be building into web and mobile applications. It is a must read for any entrepreneur building a web application.

Metrics

Cost Per Acquisition (CPA)

How much does it cost you to get a customer? It’s a simple enough calculation, how much do you spend on sales and marketing to acquire each customer. Roll up your staffing costs, your ad buys, your outbound marketing, etc.

Average Revenue Per User (ARPU)

How much revenue do users generate? How do you track it? Does it change based on segment? How do you increase it?

Churn

What percentage of your existing customer base leave every month? This is different than CPA because this is about customer satisfaction and retention. Don’t think this is important? According to April Dunford churn is a killer. “The probability of selling to an existing customer is 60-70%. The probability of selling to a new prospect is 5-20%”

Lifetime Value (LTV)

How long does a customer continue as a subscriber? Does their ARPU change over time? Do you have ways to increase their spend or reduce their churn?

These basic metrics are expanded by Dave McClure in AARRR! Startup Metrics for Pirates. Where the metrics are divided into 3 main categories:

1. Get Users (Acquisition, Referral)
2. Drive Usage (Activation, Retention)
3. Make Money (Revenue)

It seems so simple on surface, but as CEOs and startups we need to be committed to building the systems and metrics into our products. I was just floored at MeshU when I heard Dan Martell talk about the Flowtown.com Startup Immune System where they are beginning to use the lower level business performance metrics to automatically rollback design changes based on performance against the baseline. You can only start doing if you’re building on top of metrics. The idea of having automated your software deployment and sufficiently built business metric baselines that you could autoroll back poor performing changes. At Nakama, I wanted this so much. Not because I had bad developers but because we often made design decisions based on limited customer feedback and I wanted the system to protect me from my own hubris.

Metrics are good place to start. One of the best ways to understand how your company is performing is to begin measurement. Mike has done a great job