Hacker News new | past | comments | ask | show | jobs | submit login
The SaaS Metrics That Matter (sacks.substack.com)
159 points by yarapavan 41 days ago | hide | past | favorite | 23 comments

A compound monthly growth rate of “at least 15%” would be a hell of a treadmill to keep pace on. But aside from that, the target engagement figures seem aggressive. “A good metric for most SaaS startups … the typical monthly user visits the site at least two weekdays per week or 8 times per month.” “Most” SaaS startups?! I have used and loved plenty of B2B SaaS unicorns’ products that I consider the definition of successful but would never use so frequently. What business models other than social media are used so frequently?

Context is important. There are plenty of startups that fit none of these criteria but can nonetheless be considered solidly successful. The article, however, is tailored for those ones that are looking to raise a $10M Series A funding round from a top Silicon Valley VC.

Yeah. I also see now from other blog posts on their account that they like project management and note taking style apps, which, like an email client, would much more reasonably be used daily.

Those apps aren’t only daily use but have a high switching cost. So it’s a double whammy: a company that can get punters to switch into their app at a high rate (despite the cost) must be doing something right and will also have the benefit of a drag on people switching out.

Wouldn't a tool used by employees of other companies conceivably be used 5x per week for x hours per day?

That's the type of applications I build. Essentially decision making tools.

What about tools that you configure once and then don't think about again. If the thing it does is useful then that feels like a very sticky model for a B2B SAAS tool.

Not to mention if you're providing some API's you might only expect people to visit the site after they initially sign up to read the docs and then maybe not again for weeks or months.

As someone who runs a currency data API [1] I can confirm this is strongly the case - even to the point where relatively often I get messages to the effect of "we believe our company uses your service - how do we access it and can we use your logs to help us identify where in our infrastructure requests are being sent from...".

Users from clearly legitimate businesses sometimes sign up, start the billing and then just disappear. No response to emails, support tickets etc. I'm often tracking down billing teams 18 months later trying to convince them they really do use our API, and that if I can't get them to renew their billing details I'll have to eventually disable their access due to non-payment and then their infrastructure will no longer have up-to-date exchange rates! In these situations I sometimes disable access for 24 hours and then re-enable it to see if this drums up a dev I can speak to. It's quite a process but I've won some "customers for life" by saving them from downtime they would have had if I just cut them off.

[1] https://www.exchangerate-api.com

Your domain is the keywords your customers would search for. I am surprised more companies don't follow this ingenious for SEO marketing.

We registered this domain in 2010 back when there was actually something called the "exact match bonus" in SEO. Basically if your domain exactly matched the searched keywords Google just automatically gave your site a massive boost in that SERP!

I can see why they did this since there were only a few TLD's back then compared to now and thus owning the .com for a specific keyword was a legitimate signal. I can't remember exactly when this got deprecated as a ranking factor but it was very useful for the first few years!

Over the years I've accumulated the domains of failed competitors that had more punchy names and toyed with the idea of rebranding but don't think it's worth it. The SEO risk of a rebrand is substantial concern.

Your service must work so well it became invisible and forgotten. That's quite the feat. Congratulations.

Thanks - I really appreciate that! On one hand it is partly this, the uptime/reliability/LTS endpoints have been the focus of many years of work.

However, that said, I think in the majority of the actual cases I'm describing above the cause is more a combination of 1.) the surprising percentage of tech companies that appear to be in a state of continual organizational dysfunction and 2.) how API's are inherently a background type of thing like lloydatkinson wrote!

In about 1 in 10 cases I'd also say the cause is actually because of merger/acquisition - which is just always a complex and organizationally challenging process.

If you provide an API then people hitting the API are also active users of your service.

Super valuable. I made an attempt to implement something similar as a feature of a SaaS product I built, essentially taking a16z's list of metrics (like these), then building data features into the product that produced them. Generalizing it like this is way better.

A tool like this with an API would let founders and developers instrument their products with these metrics from the beginning. Even the list of user stories for the metrics features, "as a PM I would like to see the ARR/DAU/MAU etc, change over time" would be valuable.

This clarity could really advance a lot of early stage companies and save them runway time.

This is a great overview of what to measure, at least for business metrics.

(Shameless plug…) For folks wondering how to measure, and also how to track product metrics for their SaaS, I wrote a thing: https://www.tabbydata.com/book

A great idea for inbound leads. If you're doing well enough on your metrics I don't think you have to ask them if you're fundable. I am willing to bet if your numbers are terrific that they will contact you.

I built an open source app [0] (also free hosted version [1]) similar to SaaSGrid. It's especially good if you want SQL access to your modeled data. It's also integrated with Slack and Sheets so you can have your data pushed out automatically.

0 - https://github.com/mike-paper/pulse

1 - https://pulse.trypaper.io/

Should be noted, this is authored by David Sacks of PayPal/Yammer/Zenefits fame.

He was involved in Zenefits as well?

He was a previous CEO.

Wouldn’t a better metric be: value delivered/interaction? So setting up some service one which keeps delivering value seems to more valuable than something which constantly requires interaction.

Sure, but how do you measure “value delivered?” Even assuming you could measure some quantity that your product delivered (for example time the product saves the user from doing a task) you still are going to have vastly different interpretations of how “valuable“ that is.

As evidence, look no further than any thread discussing high end Apple hardware on hacker news. For High end video editors a $20,000 workstation that allows them to edit and render commercial video faster is a No brainer, because it’s a drop in the bucket in terms of total expenses and saving time is critical, allowing faster turnaround time of edits, cuts, SFx, etc. for others not so much.

For a saas example consider the superhuman email app which has a monthly subscription. For some people the value is clear and many times more than the cost. For others the fact that someone would pay for an email client every month seems insane. All because of Different interpretations of value

How to generate the file to upload in SaaS grid?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact