I agree, but I think there is a difference between metrics and having data to investigate. A metric is a value you are tracking over time, but when something changes you will likely need to do an investigation. That investigation might be the raw data, it might be some dynamic queries or it might be other metrics.
Tracking metrics because you think they might be useful in the future for diagnosing other problems doesn't make sense with modern systems as dynamic queries are so fast.
Yeah, it's hard to generalize these kinds of things since companies can operate so differently. I've seen very large companies where everyone can use the same 5 metrics and others where each team needs their own set of 5. I think the key thing is that no team is using anymore than absolutely necessary.
But it's not always clear what is "green"? For example, New Users going up can be a good thing but New Users going up by a lot can represent a fraud attack. Metrics always require some interpretation to provide value which is why we have dashboards instead of alerts.
There are some metrics that are binary good/bad and I agree that in those cases you should just have an alert.
YMMV but this is based on my experience building some of the largest analytics platforms like Flurry and Outlier.ai which were used by hundreds of thousands of companies. The only dysfunctional company I worked at was Verizon and they... don't really use metrics.
That would be true if you didn't assign a cost to each metric that you add. It was when the cost of tracking metrics plummeted that this started to become a problem, as there was no external friction in collecting more. If you assign a cost, and respect that cost, you shouldn't continuously add more.
YMMV but this was based on dozens of conversations with companies where they spent hours each week reviewing those dozens of dashboards. I was specifically talking about the metrics used by any given team, as you are right that different teams might use different metrics.
"The key is that any given person shouldn’t be using any more metrics than absolutely necessary to do their job well."
I've seen too many companies waste time on metrics theatre, tracking hundreds of metrics while knowing less and less about how the business is actually doing. The more metrics you have, the less you know.
Hi! (Founder of Outlier.ai here) You are right, our platform is designed to produce the most important insights from massive amounts of data, without requiring human supervision/configuration. It is most useful in applications when there is too much data to set up guardrails, or the teams don't know what guardrails to create. Our typical customers are very large consumer businesses who have data spread across dozens of systems and need to ensure they never miss important emerging trends or problems.
We are not an alerting or monitoring system, so I don't think you'd use us for the same applications as Orbital. The typical users of Outlier are the business users ranging from executives to business operations who want to make sure they are asking the right questions about the business.
Orbital looks like a great product, good luck in building your business!
This is a common solution to the problem of PII, but without any information on returning users I would argue that it's value as an analytics platform is limited. Few are the tools where you can grow the business without knowing the difference between a first-time and return user which is the reason cookies were invented in the first place.
However, since such businesses already need to collect personal info as part of your account creation it shouldn't be hard to build analytics on top of that existing PII. If they are already collecting PII it doesn't seem to save much to have their analytics tool avoid it?
If you are not willing to mute/ignore IMs while focusing they can be distracting and reduce productivity. However, if you aren't willing to mute/ignore I think the problem is not with instant messaging but your personal focus. Any form of communication can interrupt your workflow if you let it.
I'm disappointed they left QBE (https://en.wikipedia.org/wiki/Query_by_Example) out of their history of SQL and relations. SQL vs QBE is one of the great examples of concise technology beating over-wrought UX.
While the post makes a few good points, the underlying proposition is flawed: homework assignments are not abuse. The candidate can refuse to do the assignment if they would like. There is no forcing function on the work that would make it abuse.
In general, I'm in favor of any interview process that ensures the best people get the job. From what I've seen, homework assignments increase the likelihood the best person gets the job instead of the person who is best at interviewing. While it does take time on behalf of candidates, I'm not sure there is a better way.
Because the harassment itself becomes a limiting factor on women's ability to perform the role. You cannot have objective measurements of performance with the bias introduced by harassment.
"Differences in distributions of traits between men and women may in part explain why we don’t have 50% representation of women in tech and leadership. Discrimination to reach equal representation is unfair, divisive, and bad for business."
No, that's an over simplification. Until the statistically high rate of sexual harassment and discrimination is addressed we cannot have a valid intellectual debate on whether sex influences someone's ability as a software engineer.
No, I wasn't trying to end debate. Your position is, in contrast, an appeal to motive which is a logical fallacy.
We cannot expect the people who have been discriminated against and oppressed to stand up on their own. That is not the approach of a civil society, that is placing the entire burden on the wronged party.
That was the entire argument in the Damore essay. He made a logical jump from facts about men and women to the claim that women were less capable as software engineers.
Tracking metrics because you think they might be useful in the future for diagnosing other problems doesn't make sense with modern systems as dynamic queries are so fast.