Conversion Optimization in Practice: Baconbiz 2013 Presentation | Hacker News Reader

Conversion Optimization in Practice: Baconbiz 2013 Presentation

kalzumeus.com

26 pointssethev9 comments

'patio11 if you're reading this (or anyone else who knows): What does the "Fundamental SaaS Equation" equate to? I mean, it's obviously some measure of the "expected value" of traffic over a period... is it the "expected LTV" for a group of traffic? ie. is ARPU/churn == LTV ? (Possibly as a consequence of the infinite series implied by churn...?)

It's revenue, summed over any space of interest. (Months of life, sources of acquisition, account types, etc.)

Thanks for replying! That's what you said in the talk, too, but I don't see how it can be true unless I've misunderstood what "traffic" and "conversion rate" mean in this context:

website traffic this month x conversion rate from website = # of new users this month

# of new users this month x ARPU = revenue from new users this month

So if my churn is 5% then the fundamental SaaS equation would be equal to "20x revenue from new users this month" but that doesn't seem to factor in the existing user base at all (ie. a company with a user base which is 100x "# of new users" would see drastically different revenue that month than a company in their first or second month). Unless you're suggesting that, accounting-wise, LTV for a user should be counted as a receivable as soon as they sign up, but I doubt that's what you're saying.

Sorry for dragging on an old thread--if you'd prefer to continue via email it's in my profile. TIA for your time :)

You don't book LTV as receivables for accounting purposes, but you can plan consequential features of the business around it, like e.g. customer acquisition costs or future hiring plans.

e.g. Hypothetically assume that Appointment Reminder is currently operating at a happy place for me but doesn't have room for an employee. If I onboard 20 new paying accounts, each at a monthly price of $200 and a predicted churn rate (based on existing data) of 5% monthly, my prediction is that the business added $80k in LTV this month. This is high enough to justify adding a full-time employee.

The only place that will show on the books in the first month is as $4k of earned revenue, but we know the true state of the business is better than one would expect from just that data point. (The assumption in hiring a FTE, by the way, is that after one $80k month you continue to have other months that add at least around that much to LTV. Assuming that happens, and you can solve the cashflow issues in months 1 to 3, then you can certainly pay for that employee in an ongoing fashion.)

Very helpful, thanks Patrick.

I love this presentation. One thing to consider (slide 22) is the first step of the homepage acquisition funnel.

I have a feeling that the trial conversion is actually better than depicted. It might be worth checking to see if this "viewed page" event is including people who have already created an account, and are simply visiting the homepage to log-in.

I had the same problem with Mixpanel and wrote about it:

http://lukethomas.com/accurate-acquisition-funnels-mixpanel/

Regardless of what analytics service you use, SaaS sites should--at the very least--be segmenting guests and users. Your users behave very differently on your site compared to your guests (non-users).

If you're not segmenting analytics data by visitor type, then your metrics--conversion rate being just one of them--represent neither guests nor users. In fact, they're useless.

(If anyone needs help setting this up using Google Analytics, shoot me an email.)

Anyone else having issues with the video player? Every time I enter or exit fullscreen, it does odd things like skip ahead, rewind back, get stuck on a single frame, or (my favorite) start playing an entirely different video!

I'm on Chrome 34.0.1847.131 m (Win), FWIW.

I had similar issues in Safari 7.0.3 FWIW

Reply on news.ycombinator.com