30/10/10 | Hacker News Reader

30/10/10

avc.com

150 pointstilt17 comments

This is a great example of Fred sharing valuable info he gleans from his unique perspective as a VC who has managed to collect a lot of great Web portfolio companies.

At the same time, the super eagerness he mentions here for companies to email people, and how AWESOME that is, is different for some of us users. I can't tell you how bitterly my opinion has turned against a lot of companies I've thought were cool and whose service I thought I'd enjoy and who I entrusted with my top level personal gmail, the one that notifies me on my android just like my work email, and who I trusted to email me JUDICIOUSLY, who have ended up sending me batshit-insane quantities of email and made me very quickly hate them and the next seven generations of their prodigy forever and ever.

Yeah, at very least it should be opt in. Something like Convore which thinks it is a good idea to make daily update emails opt out rather than opt in.

Quiet a few YC companies really, it's not something you expect from the brand.

Yup. If you run opt-out rather than confirmed opt-in mailing lists, you will get blacklisted by spamhaus and the like, and kicked off of many ISPs. Opt-out, rightly or wrongly, is generally considered to be spam.

Now, I'm not saying opt-out can't be effective; I'm sure it is, else people wouldn't do it. But it does show a certain lack of respect for my time.

Progeny.

i'm not talking about unwanted email or unwanted notifications. you can turn off these emails and mobile notifications in all the major services. i do that myself. but many people like to get these emails and it does drive engagement

Thanks Fred.

I'm sure that is the case for all your companies.

I've just had nightmare experiences with some companies such as, at the worst extreme, one company who had a marketing person actually email me in response to several emails I sent begging to opt out of their frequent marketing emails to say they refused to offer me any way out of continuing to receive marketing emails, and prompting me to have to engage with upper management to the point where their management ended up profusely apologizing for the belligerence of their marketing person and ensuring me that the marketing person was being dealt with appropriately.

Am I missing something or shouldn't it be 30/10/1? 30% use it monthly, 10% use it daily, and 1% is the max number of concurrent users.

In the article the last number is 10%, but it's stated as "the max number of concurrent users of a real-time service will be 10% of the number of daily users". Isn't the number of daily users just the previous 10%? Or is that the number of users that use it within a day (which would get another ~1% from the 30% monthly users plus a little bit more)?

10% of daily users which in turn are 10% of all users is the same as 1% of all users. 10% * 10% = 1%.

Very impressive real data from Fred Wilson. In my startup it is a bit lower; 25/5/5. and how does it your startup?

I've always been weary on actual usage numbers on mobile applications. The HN crowd are probably more on the higher end of usage, but I find in general users only use a few of the well known web apps - mobile versions of their applications consistently. Approximately 30% of registrations usage of once a month is not that good. USV's portfolio companies are also some of the biggest mobile apps available now.

I really wish there was a site where these observations were all collected, common wisdom metrics. It would give us all hard goals to meet or exceed as we planned our companies. Maybe a wiki where people who wanted to share like Fred could add to the collected knowledge.

I know this is not what you meant, but Comscore has quite some data for the bigger websites.

Pretty much every company inflates their membership statistics. I wonder what the 30% figure would be if the common practice was to actually try to estimate how many individuals have signed up for the service, not merely the number of accounts.

What tools are you guys using to measure this on a web application?

Now that I revise my data, that proportion is quite accurate.

Yoda concludes bloated these numbers are.

> who have ended up sending me batshit-insane quantities of email and made me very quickly hate them

agreed, but there are also a lot of lonely people out there who without constant updates would receive no email at all. That being said, I prefer to receive notifications only when someone has interacted with a piece of content that I've shared, or has been shared about me.

Reply on news.ycombinator.com