[georgyo]

This seems really flawed.

Let's plug in numbers. This says that if a service costs 125$/month/employee and I have 1000 employees ($125000/month) I will save ~5 million a year.

However a service costing 1.5 million at such a company has other costs.

Namely that a product that costs that much almost always needs an in house support team. I have never seen the case where an expensive product also didn't need 2-3 people in just to maintain and support it. But the employee time savings is still worth it.

The next cost is much harder. In the tool we said this will save everyone 1 hour a day. However saving people time is not normally what a tool actually does. PagerDuty for example doesn't save people any time, it just sends alerts. Alerts that come from another tool that must be set up. At 40/person/month for 1000 people, that is 40,000 a month for zero "saved time." The value of pagerduty for very different.

But anyone who has been in the field has seen, that list hidden cost is that a paid service isn't normally a perfect fit. It is missing something, or your particular use case doesn't map cleanly. You then have an army of people working around the tool, saving negative time. The business might have reasons to still use that tool. But saving time is not one of them.

So seeing this tool tell me that a service which cost $2,000,000 a year is going to save me $5,000,000 feels like naive marketing nonsense. And it isn't even marketing a product.

[irjustin]

This complaint seems really flawed.

> PagerDuty for example doesn't save people any time, it just sends alerts.

Have you tried building a pagerduty in house? That is literally the time save. The "value of pagerduty" is still measured the same way as any other tool.

Do I build it custom in house? Do I pay someone else a modest amount?

That is what this tool is doing and is geared towards startups who make this decision very regularly. It's got 5 question boxes - of course, there's no way it's going to cover an enterprise consideration that needs 3 analysts to decide whether buying that SAP module is really worth $3mm/year.

Someone made a free tool and you're complaining about it as if it shouldn't exist. Very unsupportive.

Downvoted.

[georgyo]

Perhaps I was overly grumpy. But I get pulled into these discussions a lot. The company now has a very strong newly (2 years) developed "buy not build" mentality.

Half the products we have bought have had full internal teams to support. And many don't suite our needs, so we have enitre development teams building abstractions that are more complex than the product we bought so we can use the product we bought. And some of those abstractions have been in development for 2 years so no one can use the product yet.

My only point is that buying vs building is complex. When you boil that down to a tool that essentially says always buy, it causes a bit of PTSD for me.

[rightbyte]

I see this alot too. Products that are forced from above and not used, used under gun point etc. Also products that kinda don't fit in and where it would be just better to build it in house since the subset of functionaliy that is needed is quite small, and so on.

[wil421]

Sounds like your company needs to do a little more evaluation and due diligence before buy an off the shelf product. I work at “Megacorp” and it took way less than 2 years to get 20k employees and an equal number of customers onto our off the shelf ticketing platform.

[kdelok]

Even if it does boil down to that, there might still be value in ranking the savings per service you buy in. Except for a core product, it is almost always going to be the case that buying in is cheaper than building it. However, having limited resources, it's valuable to work out what tools could give you the best value for money, and which others are a more marginal call (given the uncertainty about support work etc.)

[rachelbythebay]

Do you work at Lyft?

[hedora]

> Have you tried building a pagerduty in house?

Yup. Took an afternoon. I was being cheap, so you had to give it the name of the cell phone provider along with the phone number.

Most of what pagerduty does was unnecessary for the use case, so the clone was really simple.

Also, pagerduty is harder to manage than the clone (it just had a single text file with a line per user), again, for that use case.

So, time saved is negative per user. However, the clone probably would have been harder to admin over time. It might have needed about a developer-day per month.

Most enterprise tooling I’ve seen is really about shifting work between cost centers. It’s hard to model that in a simple calculator.

[wil421]

Agreed. Pagerduty and similar notifications systems can be life savers for support people. Especially if they are integrated into a ticketing or event management system.

[falcolas]

FWIW, if you’re building a paging system, you’re doing it wrong. There’s open source projects which can do HA paging for you.

It requires a higher touch than pagerduty, but pagerduty isn’t 100% labor free either.

[whatl3y]

I feel you missed the point (or maybe that the landing page for this tool should better communicate the value and when you should use it).

If you're a relatively small company looking to use a plug-and-play SaaS product with a modest subscription fee to serve some business purpose this is a great tool to get some quick and dirty numbers on whether it would be better to use or maybe build in house or go another route.

If you're purchasing some large, highly configurable enterprise tool (with high implementation and support costs) that all 1k of your employees will use daily and you'd like to understand cost/benefit, other means of calculating this should be used (and it would likely require days to weeks of multiple people's time to make such a calculation).

[onion2k]

Let's plug in numbers. This says that if a service costs 125$/month/employee and I have 1000 employees ($125000/month) I will save ~5 million a year.

However a service costing 1.5 million at such a company has other costs.

$125/month is the price, not the cost. You need to plug in the full cost.

[falcolas]

I see this exact same thing. Let’s take Splunk into consideration. Administrating splunk, all of our feeds into splunk, and trying to keep us from overrunning our quotas is about 1 full time person, and half a full-time person worth of DevOps work yearly. That’s in addition to the subscription costs we pay to Splunk directly.

Every piece of software requires in-house support; even pagerduty and their ilk require management time to set up schedules, adding and removing people. It’s not a lot of time, about .1 to .2 of a person’s time for all of our pageable employees, but a manager’s or leader’s time is not cheap.

[hedora]

Also, splunk is considerably more difficult to use (no CLI, loses jobs, etc) and less powerful (no joins, incomplete results) than a farm of Linux log servers and some ssh tooling, such as cluster ssh, or whatever.

There’s also the question of whether the splunk log agents are more or less of a pain to administer than whatever log management they replace.

Finally, there’s the question of how the resulting reports shape people’s behavior and productivity.

If you add that all up, learning it is a waste of time for people that can code up a join in perl from muscle memory, but it saves training time for people that can’t.

In the end, every one less productive than they would be with some other tool.