Hacker News Reader

Comment by CobrastanJorji | Hacker News Reader

Anyone else think of Snowcrash and the Pizza Deliverators? In the book, the mafia-backed pizza delivery company simply guaranteed that your pizza would arrive in 30 minutes. Or else what? No, it was just a Guarantee. In the rare event they failed to deliver, they apologized in the form of outrageous vacations and such. The self-imposed penalty for failure was so great that recipients actively hope for failure.

Pretty great guarantee for the customer, all in all, although on the flip side, if your business uses Slack for critical communication in an emergency, the potential cost to you for a Slack outage might similarly be many orders of magnitude beyond the price of Slack.

View on HN

If Slack was down for 5 minutes in a month that will break the 99.99% SLA. If you pay $6.67/mo for Slack, 6.67 * (0.01%) * 100 = 0.07

Yea I'll rather have Slack for that 5 minutes than save my company 7 cents.

If your company is only paying Slack $6.67 a month, you've either negotiated for an amazing discount or I really hope your chatbots are fun to talk to.

But I agree with your assessment: the purpose of the 100x is almost certainly to make it less offensive to receive an SLA credit. The only thing worse than having downtime is getting a credit that doesn't even round up to a dollar.

I'm just using the per user cost here. $6.67/user/mo is the standard plan price if you pay annually.

Most companies tend to have more than one employee - especially when they decide to buy an employee chat software.

As a matter of fact, the median company size in the United States is 1 - almost 80% of businesses have no employees at all!

But I agree that most of those 0- and 1-person businesses are probably not Slack customers ;)

7 cents per employee per month isn't exactly a generous discount either. Even if your chat is down for half the work day (4 hours), you're looking at $3.55.

I mean, it's better than nothing, but let's not pretend it's compensation for lost productivity either.

My thoughts are this: even at 100x, the service credit is probably not that significant to the recipient. But the amount at 100x is significant to Slack. As the recipient, I don’t think too much about the money I receive back, but I know those credits in aggregate are quite painful to Slack (and a lot more painful than they could have otherwise justified) so I know they are taking the issue seriously and will work hard to prevent future outages, which is worth a lot more to me than the service credit.

I'm using unit cost here so it's easier to see how I personally feel about it. Just multiply the number by however many employees you want to imagine.

And if you have 100,000 employees on slack, that compensation will be $7,000. For a company of that size this is not really noticeable. Refunds don't really matter, I guess most clients would be happier if that money was invested in increasing future reliability.

Pretty sure most companies have more than one user

The SLA only applies to the Plus plans.

Peter's point still holds. If the company has 100 employees, that is $7! For 10,000 employees, it comes to $700!

His point is that it isn't really significant. How can "catching" him on the single employee assumption be the response.

The number of companies for whom Chat SAAS is a major expense is probably pretty small. I think the refund is more important for how it impacts Slack than how it impacts its customers. As long as the refund is substantial enough to significantly impact Slack's bottom line, then I'm more inclined to believe that they are going to make a serious effort to avoid paying it in the future. That said, I'm not sure that refunding at this level represents a sufficient incentive.

7$ for 5 minutes of downtime in a month? That seems good to me. 1 hour would be a better example I guess, 84$ for an hour of downtime, that seems to be low for the amount of work that "may" have been lost, but then, that's weird to depend on a chat software that much.

I don't see what's weird about that -- before companies relied on chat, they relied on email -- in my company email can go down and people hardly notice, but a Slack outage is immediately met with cries of "Hey, are you having trouble with Slack!?".

Before companies relied on email, they relied on phones. My company doesn't even provide desk phones for most staff, only Sales and other staff that need to make a lot of calls have them.

So what's so weird about companies relying on Slack?

It is worth noting the SLA only applies to the Plus and Enterprise levels, so it would $12.50 / user. Still only 12.5 cents per 0.01% downtime per user. Doesn't really change your point though.

They were down for 53 minutes by their reckon and 115 minutes if the OG tweet times are right. Rough split the difference and say 65 minutes downtime, the refund would be ~ $62.50 for a 20 person team bill per month at current rates. That is not too shabby for being out of service for 1 hour during North American primetime. If they were out for 4 hrs 30 mins, it would have been a full refund for the month for everyone!

At 8 hrs outage/month, Slack would be completely free.

And that's a great way to test that you have backup comms systems working.

At 8 hrs outage per month, slack would have been out of business 4 hours ago.

In Snow Crash the boss would show up if your pizza was late. Nobody wanted to be responsible for the boss having to show up since the pizzeria was owned by the Mob.

Reply on news.ycombinator.com