The source of this differentiated supply is individual hosts. Airbnb basically created this supply that mostly didn't exist before, especially at this scale. In 2018, Chesky mentioned 70% of properties listed on Airbnb was unique. Some estimates indicate that number was likely as high as 80% in 2020. While unique supply is certainly a competitive advantage, there are more nuances to it that may dampen the extent of this advantage which I'll explain later.

Even though individual hosts dominate Airbnb platform today, over time it seems somewhat inevitable that professional hosts will become majority of bookings mix. The real question or debate is probably around time i.e. whether it takes 5, 10, or 15 years. Why is this almost inevitable?

If you look at occupancy rate at Airbnb, it's been consistently ~50-55 days per year or ~15%. For context, hotels such as Marriott reported 73.7% occupancy rate in 2019 (pre-pandemic). Airbnb's occupancy rate reflects the current dominance of individual hosts who mostly make their apartments available for rent for a few weeks in a year. For Airbnb to maintain its high growth rate, it is essential to grow its listings and/or occupancy rate. Professional hosts, who often manage anywhere between tens to thousands of properties and are much more driven to keep occupancy rate high, can be far more efficient way to maintain the growth momentum.

oftware continues to grow well, but it's not growing as quickly as some of our other verticals. And I think we attribute a lot of that to the fact that software is an area where we've historically had early adopters come in and really kind of drive improvements in their go-to-market motions. And many of these other verticals are more lightly penetrated and almost certain that software has been reviewing more of their spend in the last 3 to 6 months.

transportation logistics, financial services and insurance,

whether that's Intent, Chorus, Engage, MarketingOS, the OperationsOS and the data as a service and the TalentOS as well.

in order to give the enterprise a huge amount of flexibility in how each interconnect was performed. SD-WAN could provide a centralized control

I think close to that, a hybrid approach like that would be wise. I think they've got to put a whole lot more fuel on the fire with Flex and Segment . They have to. And they have to differentiate themselves. They have to steal more customers away. They can't keep getting beat out on little one-on-one battles with the Vonages of the world because Vonage has a better CCaaS. They can't beaten out by Infobip because Infobip has gateways.

They're going to have to go in and they're going to have to spend more dough and do be more active about moving more of the SMS messaging into OTT. They're going to have to do that in APAC. They're going to have to establish relationships with WeChat, LINE and Kakao and do the same thing that they just did over the last 1.5 years or so with WhatsApp because that worked.

If they shut them off, right, because they have frenemies right now, companies like Infobip and Sinch use Segment. So they actually use Segment, so they could just be like, hey, nice having you, but we're not re-upping you.

“80 or 90 important models will carry about 90% of the freight in making you a worldly‑wise person.”

(1) Hanlon’s Razor — “Never attribute to malice that which is adequately explained by carelessness.” (related: fundamental attribution error — “ the tendency for people to place an undue emphasis on internal characteristics of the agent (character or intention), rather than external factors, in explaining another person’s behavior in a given situation.”)

(1) Cognitive Biases — “Tendencies to think in certain ways that can lead to systematic deviations from a standard of rationality or good judgments.” (See list of cognitive biases)

Long term (with the caveat that the markets work well and the competitors are rational), a niche for data can be dominated by 1 or 2 players. That dominance does not give these players pricing power. In fact, they actually might have negative pricing power (one of the ways a company may continue to dominate a data market is by lowering its price to make it harder for rivals to compete).

the data company can dominate by just having the correct facts and having an easy way to deliver those facts (APIs, queryability, self-serve, and integrations become very important).

Those companies include core tools (Databricks, Cloudera), middleware (LiveRamp, Plaid), BI (Tableau, Looker), data processing (Snowflake), log processing (Splunk), and many, many, many more. (note: as a reminder about the power of these tools … while I was writing this post, both Tableau and Looker were acquired for a total price for almost $20 billion!)

THE SECOND fundamental truth of data business models is this: whoever controls the data, captures the value. Intermediaries get squeezed.

THE FIRST fundamental truth of data business models is this: it’s all about the data.Successful data businesses are all built around a unique or proprietary data asset. There are a few ways to build such an asset:

And at the end, my data lands in the solution, right, even though I'm not even aware of this, that it happened. That's the issue. And one way of kind of solving that issue is requesting that user to include some sort of privacy notice in these messages, right, potentially, at the bottom, most people don't read that part on the email I assume. But this is the way how we can target this issue, to include some sort of short privacy notice and a link to privacy policy there.

Right. So basically two general rules. The first rule applies to whenever you collect the data straight from the data subject, right? So let's say, this is just an example. You go into a bank, right, you sign the agreement and they inform you at this stage because they collect the data straight from you. The other rule is when you collect data from somewhere else. So for example, you go to the database, you take data from there. And you have no touchpoint at this stage with the data subject. And speaking of GDPR, it requires data controller.

So those are basically two rules how you should do it. And as you said, ZoomInfo says that they inform all the members of the platform, right, they send an email. If they do, that's fine. If they don't, that's not only their issue because everyone using the data they collected illegally, in fact, in this scenario, right, so let's say, those data are already poisoned, right, because they were improperly collected, without fulfilling the information obligation. So that's the biggest issue.

Certainly. So I think the helpful way to frame any company like this when they're touching this type of information is first to distinguish, like what is the question you're really trying to ask here. So I want to play that back because I think it's really important.It's less a question of did they access the data properly, right? Because you've seen the Supreme Court has recently ruled that if it's publicly available, LinkedIn tried to say that their public profiles if you were scraping it that, that wasn't appropriate access, you're actually violating the hacking law.And what was effectively ruled there is like, no, it's appropriate access. I think the important question you're asking is a separate one, which is, okay, maybe they're getting the data and nobody is sitting there saying that they hacked in or that they had the crawling method of publicly accessible data is wrong. But like are they gathering data for certain uses and then are they applying it? Are they activating that data? Are they selling it in violation of any use limitations or tripping over any privacy regulations?And the two ways to think about it are: one, when they're crawling public data, which a lot of these contact data platforms do; and then two, when they're augmenting that with commercially acquired data. That's a generic term, but it typically means they're buying data from a data broker or a list provider.And in both cases, I firmly believe that one of the most important things to do at a really simple level is think about the context in which the data was initially provided and then the purposes for which that data is being used, and we're going to come back to that towards the end. But I think that's a really important way to think about it over the long term.In the short term, the reality is, is that they're gathering that data, somebody is gathering that data. And in the case of ZoomInfo , let's use a very simple example where somebody signs up for some information that's being provided to them on the web, and that information provider, maybe they're providing some rankings on software, things like that, they in their privacy policy, they say, "Hey, we may actually turn around and sell this to other people." Or and/or this data may be publicly available and it could be used for other things, right?So somebody through that privacy policy is being made aware of that. The information, if it's not sold to ZoomInfo , when ZoomInfo is doing that crawling, CCPA, the California privacy law that's currently on the books, is not so clear about whether that constitutes a sale.And when we talk about what I mean it's not so clear that it constitutes a sale, think about it, that individual's contact information was being provided to a first-party data owner, maybe they're providing a webinar, rankings of software, whatever it is, and then that data is being used by ZoomInfo for other purposes. Without ZoomInfo paying for it, CCPA was sometimes hard to read clear whether that constituted a sale.And so the big difference happening on January 1 is that CPRA, the successor to California's CCPA, is very specific in that it separates out selling from sharing. And they say, "Look, if you're sharing the data, individuals have a right to opt out of that sharing." And in a lot of cases, they'll require affirmative action of the user. The user has to take some kind of action.And so the key questions to ask if I'm sitting where an investor is sitting or where anybody is sitting that to say, does this affect the economics of ZoomInfo 's business? I think it's ultimately one of the important questions asked here is ultimately how much is ZoomInfo going to have to pare back the data that they're gathering?Either because when they're crawling publicly available data, they have to care for certain use limitations and cut back on it. Or because when they're purchasing the data from these third-party data sources that they used to buy from and still buy from today, they've got to go back and ask them, "Hey, is this all being done with the proper use limitations?"

Well, think about it from the individual's contact information, they are providing it to somebody who typically is not ZoomInfo , okay? They will provide it there. It will be gathered, shared with ZoomInfo in some active or passive way.ZoomInfo is going to do two things that are applicable to your question today. They're going to one, store it in a profile for the purposes of advertising, marketing, messaging to that same individual, but on a different site than the site where the individual provided their contact information.So here's an example. John Doe works at an e-commerce platform. They run all of e-commerce there. And they're interested in some payment security platform, right? So there's a lot of e-books out there about that. And maybe there's a website that's gathered all these e-books. So you go and you download a couple of them.It doesn't even matter which vendor, right? It's just the mere fact that you were interested in payment security vendors. While a payment security vendor may come along to Zoom Info and say, "Hey, I just really like to find people who have an e-commerce title, right? Maybe you provided that when you were looking at that, looking up that collection of e-books.So when you provide that title information, you're providing it on one site, then ZoomInfo is then using it in a different website for reaching you. Maybe you're just going on to link it, and they're running a LinkedIn ad there, right? Maybe somebody comes along and buys from ZoomInfo some data and then reaches you when you're on a business journal.So that's a different website. But also importantly, back to this high-level framework I was talking about before, it's a different context. And that gets a little fuzzier because if you listen to the Federal Trade Commission and some of the other data privacy regimes, they really ultimately get very concerned about people providing user data, personal data in one context, they're being used for another.It's sometimes really hard to have that different context be codified in these regulations. And so what they tend to do, even when they're talking about cross-context behavioral advertising, they're really talking about distinctly different branded, say, websites.

I can't identify you personally, but I can gather a lot of information about you. I can store it, I can store that behavior. And then in a different context on a different website, I can use it to reach out with a different message. So that's cross-context. That's sometimes called multisite retargeting. That's exactly what CPRA is most directly impacting here when you think about the types of businesses we're talking about with ZoomInfo .

Yeah. Yes, but data vendors, man, it's like using plug and play. They might not be as good, but there is no real deep moat in that space. If Zoom didn't exist tomorrow, we would just go find another data vendor or a meshing of data vendors in order to fill the gap we need.

As long as they keep consolidating their competitors they've picked up, they'll be fine.

So they consolidated Datanyze , they consolidated ZoomInfo and DiscoverOrg. If they go out and buy Clearbit and some of these other ones when they start coming up, then I think they'll be in a good place. I don't know what the moat is I guess is what I'll tell you.This is a really easy business for anybody to break into. Data businesses, a lot of these little companies, they get to 10 million in ARR and they get scooped up by the big guy. It doesn't really add much more value to the big company unless it gives them a whole new dataset.So my worry for I guess investing in Zoom or DiscoverOrg, or this conglomerate or whatever it is now, is do they have to keep acquiring in order to stay ahead? Not acquiring new features or whatnot, but just acquiring new datasets. Is there enough of a moat to make this like why is this the one. What does LinkedIn do?Now, I don't think you have to worry about that as much, but it's there. That's a wildcard. Then what happens with GDPR and CCPA? As that starts to kick up, what other places adopt that and how does that affect how people use the data or secure the data?