34 Comments

at some point every friday afternoon everyone on my team spends like 10 minutes talking about your latest blog post

Expand full comment

Tell your team (and investors) I'm sorry.

Expand full comment

Yet another long-time data company preaching the faith:

"We empower businesses to realize transformative outcomes by bringing their data and AI to life. When properly unlocked, data becomes a living and trusted asset that's democratized across the organization."

TO LIFE I SAY! TO LIFE!!!

https://youtube.com/clip/UgkxggSu3VV65rcBkGc8DLgveYluORVgMSiM?si=5wepS-rctZhpH8-X

Expand full comment

I'm guessing that's Informatica? But the thing is, my first thought was it's entirely possible that's marketing copy from Mode that I might've written.

Expand full comment

It is indeed.

Benn, you're a much better writer than that!

Expand full comment

But also, maybe they're the ones who are on to something

https://photos.app.goo.gl/BD5PG9uojpWypaAj7

Expand full comment

Not because of that copy!

Expand full comment

You have talked about accountants and data in the past - and I’m thinking the accountants view of data is more facts than faith. I wonder if there will be a swing back toward data is accountability, data is a clear view of what did or did not get done. Did you make budget or turn a profit? Did you hit your sales number? It’s easy to brush by things like that as table stakes and of course everyone is doing that stuff well. Then announce everyone should be doing predictive analytics and AI... but I’m here to say most everyone does not seem to have it together with basic metrics and goals - atleast from my personal experience.

Expand full comment

Yeah, very much agreed. But you can't build a religion around accounting, which sounds a lot less fun.

Expand full comment

Ya - at least not a very popular American religion. Just following a bunch of rules as your religion has gone out of style it seems. But I suppose even if that was your religion you have some faith in following the rules will get you something.

Expand full comment

Real data work is like being a Catholic - it's all rules and guilt and disappointment. But everyone wants to go to some non-denominational church that's mostly there for vibes.

Expand full comment

I mean - I believe there are other options too - like genuine faith - but that’s about right for a lot of people unfortunately.

Expand full comment

Hahaha. Don’t think I can disagree with that.

Expand full comment

Benn as Naval says “ if u live long u become a philisoper” one more data tag line from my company “ Human data science company” …… good read mate

Expand full comment

I guess there's a corollary to that for data companies; if you survive long enough, you become a religion.

Expand full comment

Data is just information. We're naturally drawn to faith that promises to organize all the information we perceive about our reality into a coherent narrative. One that will allow us to navigate to success or salvation. We are selling the same promise to organizations. We will transform the noise of your reality into meaningful narratives that let you navigate towards growth / profit.

Expand full comment

So yeah, I agree that that's the pitch to get people in the church. But I don't understand why people believe it. At some level, sure, it's obvious - we should know stuff; data is information; information helps us know stuff. But there are lots of other sources of information, and data does seem to be the one we ascribe some special, almost spiritual, significance to, despite a lot of companies actually struggling to use it in that way.

Like, I agree, in theory, data *could* help you deftly navigate towards growth, profit, etc. In practice, it's kind of a struggle? And yet, we keep the faith.

Expand full comment

People go to church for many reasons; despite a lot of reasons why the proposed truth it's peddling is bogus. You mention one of them in this blogpost - community. "We are not truth-seeking animals, we're social animals" to paraphrase. We - be it as individuals or organizations - want to believe what everyone is believing because it's cold and dangerous out in the cold (even if that's where there's more "Truth"). The church is warm.

And at least in my view capital-D Data does offer the best method of navigating towards our goals. In terms of cybernetics - improving our sensor should lead to better control and create virtuous feedback loops that help us orient and move towards our goals.

What's your take? What is the right amount of belief in Data?

Expand full comment

Yeah, I agree there's a lot of mostly rational reasons to buy into it. There's a like-mindedness among "data people" that's appealing; people share roughly similar value structures; more cynically, people like feeling "saved," and believing that you've found the Truth - be it your religion or your rationality - is comforting. And that's useful, even if you're just worshipping a teapot orbiting the sun.

To your question of whether or not that belief is right, I don't know. My boring answer is that it depends? There are some businesses and problems where it's great, and some where it's not that useful.

Expand full comment

It's useful to direct your energies and adapt your language to the belief structure you find yourself in. On top of that, a healthy dose of doubt can keep you open to doing things differently or exploring beyond "the edge of the world".

I like your answer. I hope to get a better understanding of what businesses and what type of problems most benefit from this rational Data ideology and which ones benefit from being more wild & free.

Expand full comment

That's an interesting thought I hadn't considered - that, even if you aren't convinced it's true, there is some value in "believing" in some form of faith. It gives you a belief structure, community, a lens to see the world. I could actually see the data thing having some of the same benefits, where it's all true-ish, but there's a good bit of personal value in buying into the faith.

Expand full comment

TL;DR

I can be selfish but that's an aside. People built lots of things. Some of them were good and markets are different now. Weird social things also happened - that is another aside.

Now people can build better things because time is linear and now is later than then. In theory, we can only improve. Something good will be built, we just don't know what yet.

How'd I do?

Expand full comment

I mean, it's not wrong.

Expand full comment

we need Ryan Radia on that Divorce stat

Expand full comment

I'm not even sure I even figured out what the divorce rate actually is. Half the time, it seems like it was # of annual divorces / # of annual marriages, which I guess sorta works? And also, as an aside, there's no way someone hasn't made some horrific double counting error on this one somewhere.

Expand full comment

most public stats are like that.

I think most of the time the divorce rate is (divorces in a year)/ (married people in a year+divorces) *1000

Expand full comment

I believe the marriages is taken from an end of year or beginning of the year stock, whereas the the divorces are an event .

Expand full comment

I believe it is usually just done with women to prevent double counting. And yeah, it’s a convention to calculate many types of population statistics as “per thousand [humans of some type]”

Expand full comment

Per 1000 makes sense; only using women seems...problematic?; and count/stock seems mostly fine, though then it becomes, like, the chances of getting a divorce *this* year instead of will you ever get a divorce? Which probably makes more sense as a long-term demographic trend, but is less useful when comparing real people to The Bachelor people.

Expand full comment

yeah I the women only thing seems like it turns same sex couples into a real counting nightmare.

The inference (and it is an inference) of "what's the p(getting divorced | getting married)?" ever is probably difficult from any source. It's very easy to go from monthly divorce rate to how many months should the average marriage last, but you then you are effectively going to have to compare the relative likelihood of divorce happening before mortality.

Expand full comment

five year divorce rate or ten year divorce rate is probably fairly easy though

Expand full comment

I can't even begin to make sense of that formula. So like, if two people were in the world and they got married and divorced in the same year, it'd be...

1 / (2?... + 1?) * 1000? So...1000/3?

Or I guess it's married people at the end of the year? So if there were four people and two pairs of married ones and two of them got divorced it'd be...

1 / (2 + 1) * 1000? Which is still the same thing?

Expand full comment

deeply satisfying and comedic

Expand full comment