Jim Lanzone has spent his entire career betting on companies the internet forgot about and making them matter again. CEO of Ask.com, Tinder, and now Yahoo. Eric sat down with him to find out how he pulled off the turnaround nobody thought was possible, what single decision handed Google the internet, and why Yahoo is building its own AI search product for the first time in over a decade.They get into the $44 billion Microsoft deal Yahoo walked away from, why AI chatbots are failing publishers and what needs to change, how Yahoo Scout is trying to do search differently, and what 700 million users actually looks like in 2026.If you think Yahoo is a relic, this conversation will change your mind.
00:00:00
Those of us you know who ran search engines, we'd be on stage
00:00:03
and, and and talking. But in the audience were all the
00:00:06
spammers, all the. Whole SEO we take advantage.
00:00:09
Of the whole SEO community and if you went outside or to the
00:00:12
bar afterwards, everyone's together and that game of trying
00:00:16
to send traffic for their own purposes, that's not new.
00:00:19
Why is Yahoo Finance still #1 why is Yahoo Sports top two?
00:00:23
Why do so many people want aggregated news?
00:00:25
And like, what can we do to lean into that?
00:00:26
Why is Yahoo Mail still so popular?
00:00:28
And it it the spine of that book is is always goes back.
00:00:31
What is it? What's?
00:00:32
The answer It always goes back to Jerry and David's original
00:00:35
founding of the company. They were trying to be the guide
00:00:37
to the Internet, but every now and then that is.
00:00:38
The norm that is contrary. To OK, but think about why that
00:00:42
happened. Everyone who was involved was
00:00:44
not a traditional search product manager or engineer.
00:00:47
They all came out of academia and so the the UI, not
00:00:51
surprisingly, looks like an academic research paper.
00:00:54
My guest today has been CEO of ask.com, CEO of Tinder, and now
00:00:58
CEO of Yahoo. If you're sensing a theme, it's
00:01:00
that Jim Lanzone has spent his entire career betting on
00:01:03
companies the Internet forgot about and making them matter
00:01:07
again. The question is, can he do it
00:01:08
one more time with guy closing in from every direction?
00:01:12
Here's my conversation with Jim Lanzone.
00:01:22
Search wars, the streaming wars and the search wars again.
00:01:25
And I'm sure there are 50 other online wars you've been engaged
00:01:29
in. You're obviously running Yahoo.
00:01:33
Some people remember it from there.
00:01:35
You know, maybe me. My college e-mail I think was
00:01:38
Elfman. I won't say the exact numbers,
00:01:41
Elfman something back in sort of a we'll go find AD and D type
00:01:44
era of Eric Newcomer, but obviously still a huge Center
00:01:49
for traffic. You're obviously doing a lot to
00:01:52
stay current in the AI era, like where is just give us the
00:01:56
snapshot of sort of Yahoo in terms of who's using it today.
00:02:00
Yeah, I think it's it's surprising people for some
00:02:02
reason, even though, you know, Google is 28 years old and
00:02:05
Amazon's also 30 years old. eBay, I think Twitter just had
00:02:08
its 20th anniversary. Yeah.
00:02:10
It's funny that Yahoo, probably just because the history of it
00:02:13
like does kind of get talked about as being around a long
00:02:16
time you. Know it defined 1.
00:02:19
Era, it has had a lot of ups and downs and that's part of the
00:02:22
story. And and you know, we can talk
00:02:24
about some of that. But the interesting thing about
00:02:26
Yahoo is for all the corporate difficulties, you know, through
00:02:29
time as a public company selling to Verizon, Then we bought it
00:02:32
and spun it back out at the end 2021 which.
00:02:35
Seems like a good deal. Apollo got it for around 5
00:02:39
billion after all that, so it is still today.
00:02:44
Now Fast forward to 2026, in the last comScore rankings, we were
00:02:48
third in the United States until traffic.
00:02:50
We're almost 250 million users in the US, which is almost 90%
00:02:54
of the US Interestingly, about 50% of our audience stays Gen.
00:03:00
Z and Millennial. I think when you cover 90% of
00:03:03
the US, you have to be that cover that broad of a group and
00:03:08
that's actually our fastest growing group and things like
00:03:10
Yahoo Mail, which you might associate as being an older
00:03:12
product for the Elf man and others.
00:03:16
And it's about 700 million, you know, worldwide users and that
00:03:19
that's after we've shut we actually got out of a lot of
00:03:22
countries when I first got here and we were the turn around
00:03:24
team. There were a lot of things that
00:03:25
we had to do to kind of get the company on the right foot.
00:03:28
So we kind of got out of some countries that that weren't that
00:03:31
didn't make sense. So it is still that big and the
00:03:35
way we get to that size is different than a lot of you know
00:03:38
there are a lot of companies that are one product companies,
00:03:41
right you got. Finance, you've got sort of this
00:03:43
and. We are a conglomerate.
00:03:44
They're, they're all Yahoo branded.
00:03:46
And we, we did have some brands like AOL that, that were not
00:03:49
Yahoo branded that we've sold along the way, but we are now
00:03:52
all Yahoo branded. And yeah, finance, sports, mail,
00:03:56
news is still top three things like weather are still, you
00:04:00
know, top five and still still very big properties.
00:04:03
And then search, I think is, is still surprised that in the US
00:04:07
we're still the third ranked search engine.
00:04:08
Now that's that's far away from from Google, but but still a lot
00:04:12
of traffic in a very, you know, lucrative category.
00:04:15
So we're all those things together kind of add up that the
00:04:17
Yahoo brand itself doesn't mean one, there's not like, you know,
00:04:20
just one property that yahoo.com, which we've kind of
00:04:24
been working on over the past few years is a hub for all those
00:04:27
things. But you know, again, there are
00:04:28
many companies that just do one thing, and that's not us.
00:04:32
Is and there are lots of pieces of it that I want to talk about
00:04:35
you used to own tech crunch. You touch the news, obviously as
00:04:38
a news person, it's interesting to me.
00:04:40
I but you know, you Yahoo obviously fought with Google
00:04:44
through search, made some, I guess failed partnership
00:04:48
decisions. You've criticized them yourself.
00:04:50
What was what was the key mistake?
00:04:51
It's sort of outsourced search, right?
00:04:53
Well, it's the interesting thing is Yahoo has always talked about
00:04:55
as having lost search to Google and, and kind of shit on for
00:04:59
that and talked about what, what you shouldn't do strategically,
00:05:01
but that's not actually what happened.
00:05:03
And again, that's 26 years since this happened.
00:05:06
And, and it was before many generations of Yahoo, let alone
00:05:10
mine. But from the beginning, 1995, it
00:05:13
always Yahoo always outsourced search, interestingly, because I
00:05:17
think they helped popularize search, but that search box was
00:05:21
outsourced through enterprise deals.
00:05:23
So I can't remember who it initially was, but it became ink
00:05:26
to me. And then during the crash
00:05:29
period, during the first web, you know, one point O crash in
00:05:33
June of 2000, which is kind of the big month that it really
00:05:36
felt everything fell apart for the industry.
00:05:39
They switched to Google because it was the best search engine
00:05:42
and that became the enterprise solution for Yahoo Search.
00:05:46
The big mistake that was made is that in order to save money and
00:05:49
get a better deal, they gave Google a link back to Google on
00:05:54
every search results page. That was the Google logo.
00:05:57
So if you did a search on Yahoo and you clicked on that logo, it
00:06:00
would do the search on Google. And so people just they're.
00:06:03
Training people just go. Oh my, I've said it's it's
00:06:06
almost like Chacha PT or clawed having a logo on every Google
00:06:10
search results page now you know, they would.
00:06:13
So that's that, that kicked off, you know, I think, you know, a
00:06:18
couple decades of trying to figure out, you know, how to
00:06:20
make up for that because Google, you know, search became, you
00:06:23
know, the, the best revenue model that you could possibly
00:06:25
have. But in June of 2000, there
00:06:27
wasn't one. Right.
00:06:28
And you can see the philosophy. It's like we link out to news
00:06:30
organizations where helpful. We we get people to go to the
00:06:34
place we trust where you know, yeah, where the portal.
00:06:37
But obviously in retrospect, a huge strategic error.
00:06:41
So, you know, you're not even thinking probably about the
00:06:44
company at the time, but now, you know, you run the company
00:06:47
that made that decision. So now we enter the chatbot era.
00:06:51
Like do you think first of all, does the chat bot replace the
00:06:56
Internet or where, how much of sort of Internet activity do you
00:07:00
think will, you know in a decade flow through me talking to a
00:07:04
sort of an AI chat bot? Yeah, I mean, I'd say still
00:07:10
somewhat unclear. It's, it's a bit of a misnomer
00:07:14
to say that finally now going beyond 10 blue links in my, my
00:07:18
first run through search, you know, I spent my first ten years
00:07:20
in or so in search. That was when we really went
00:07:23
beyond 10 blue links. It was literally 10 links.
00:07:26
Alta Vista kind of popularized that UI.
00:07:28
Google picked it up from there. And then in the mid 2000s, we
00:07:32
all started building structured data search into the page.
00:07:35
So if you think about like. You're getting answers.
00:07:37
Whether lyrics, sports scores, you know, multimedia, these
00:07:42
things all started to have local information maps.
00:07:45
They all started happening over time.
00:07:48
So it's interesting now that area is popular, you know on
00:07:53
Google is populated by AI mode. So now you're kind of getting
00:07:57
chatbot responses, which are their own flavor of search.
00:08:00
They're not, they don't replace all of search.
00:08:03
And I think this next step of a gentic, which is still very
00:08:07
early, I think thinking about the the search box as the
00:08:12
interface for doing anything probably does make a lot of
00:08:16
sense. Interestingly, because that's
00:08:17
where everything started. If you think about how Yahoo
00:08:19
started in 1995, it was a directory that then led to news
00:08:23
and mail and fancy sports and other things coming into that
00:08:26
becoming the portal. But it started with the search
00:08:28
box. And to think that all the way
00:08:31
through now, 30 years later, that search box is about to
00:08:35
become the Iron Man suit and and power everything that you might
00:08:38
be able to do to do is really interesting.
00:08:41
I do, I do think it tends to go that way.
00:08:43
And you see that now where WhatsApp has a chatbot, you
00:08:47
know, built in, Snap has it built in.
00:08:50
So it's that is starting to happen.
00:08:52
And what that turns into, I think is still to be written to.
00:08:55
Defend the portal. My problem with chat bots is
00:08:58
that I need to have an idea of what I want to do and it's sort
00:09:01
of like I really need to have a plan and really sort of pursue
00:09:05
something. There's value to be like, oh,
00:09:07
what's going on? Like give me, give me sort of a
00:09:10
picture of it. Like do you think sort of the, I
00:09:13
mean, you're, you're sort of playing both in that you're
00:09:15
making sure you're in a good position on chat bots and you
00:09:18
obviously have the the key portal strategy.
00:09:20
But do you think the regular consumer is going to want this
00:09:24
chat bot all the time for all their sort of like information
00:09:27
needs? Or is that is it too much work
00:09:29
to ask the news consumer I. Think people in our industry
00:09:33
can, especially everyone who's building all night now, you
00:09:37
know, with all these new tools that are coming, you know, are
00:09:41
available, do forget how much the industry itself is, is not
00:09:46
the same as your average user out on the street.
00:09:49
So those people do tend to want things done for them.
00:09:51
They want things that are free. They don't want to put in the
00:09:55
extra effort. I mean, if I go back to web one
00:09:57
point O search, my favorite was that 85% of users if you
00:10:02
surveyed them said they wanted advanced search, which you
00:10:05
probably don't remember what that is.
00:10:06
It was boolean. Search I remember.
00:10:09
One less than 1% used it. They they said they wanted but
00:10:12
they want the power. It didn't actually they, they
00:10:14
want and we've all put in a lot of effort in that day and age
00:10:18
into thinking about how to get, you know, them to construct
00:10:22
queries a certain way. And that all went away.
00:10:24
It just became go to the search box, spit it out.
00:10:27
And it's up to us on the other side of that box to deal with
00:10:30
it. And I do think, especially when
00:10:32
it comes to search, that that's where it's going to head.
00:10:34
And and we could talk about it, but we, you know, you know, a
00:10:38
month and a half ago launched our own AI search product called
00:10:42
Yahoo Scout. And to your point in the portal,
00:10:44
we also just launched last week the next generation of what used
00:10:50
to be known as my Yahoo, which was very popular Silk high value
00:10:53
product, which is your own personalized version of the
00:10:56
Yahoo portal. I feel like I had all these
00:10:58
different colors. Yeah.
00:10:59
And people loved it and they could move things around on the
00:11:00
page and but that was done in a very rudimentary, you know, old
00:11:05
school way. But it was popular.
00:11:07
We had to shut that down when I got here because the the
00:11:11
platform it was built on was so old that to upgrade it was just
00:11:14
going to be just incredibly expensive.
00:11:17
And we just decided to build it over again, knowing that AI was
00:11:20
on the way. So we just last week launched my
00:11:23
Scout, which looks a lot like the old my Yahoo, but it is it
00:11:26
is built by and on top of our AI search platform.
00:11:31
So it it automatically builds for you on on the page and just
00:11:37
by querying, you know, querying scout, you can actually create
00:11:41
your own categories. So anything you want can now
00:11:45
become a tile for your own personalized portal on the page.
00:11:48
Right. So it's not the chatbot can help
00:11:50
you sort of have defaults and you can already see sometimes
00:11:52
with Chachi BT, they're like suggesting questions you can ask
00:11:56
like it's clear that all that's yeah, that's different than just
00:11:59
like a portal you go to, but well.
00:12:01
And also they're, they are limited to date by what the, the
00:12:04
thing they know about you is your search history.
00:12:09
And so Pulse, which is their first product was very much
00:12:11
based on the searches you've been doing.
00:12:14
Oh, do you want to know how to become a better CEO of Yahoo?
00:12:16
How do you about these, these health issues you've asked me
00:12:19
about, you know, whatever these things are going on with your
00:12:21
kids, like whatever you've been asking is kind of what they
00:12:23
suggest up in their portal. We do have a built in advantage
00:12:26
in that if if you are a logged in Yahoo user and over 70% of
00:12:30
our users are logged in on any given day, we start to populate
00:12:34
it automatically with Yahoo Finance.
00:12:35
You know, with stock quotes, with sports scores, with
00:12:37
headlines, with your mail, all those things, your, your
00:12:40
location, your weather, they all automatically populate and then
00:12:43
you can add categories onto it from from there.
00:12:47
All right, so you're, you're preparing for new fronts right
00:12:50
now, which we are. You know what, you, you, I, I
00:12:53
never covered like the media streaming wars.
00:12:55
I, it was very important, I think in the, the peak streaming
00:12:59
days, but it's, it's a moment to talk to advertisers.
00:13:02
I, I imagine it's especially interesting with open AI mean
00:13:05
like we want to get into advertising your sort of, you
00:13:08
know, have more and more of an AI pitch.
00:13:11
Like what is the dynamic in new fronts broadly today?
00:13:14
Like what do you think the key themes that people are grabbing
00:13:16
onto? And and then we can talk about
00:13:18
Yahoo specifically. I mean, I think different for
00:13:20
everybody. The way the way the new fronts
00:13:22
started many years ago was to give online streamers and I used
00:13:27
to work at CBS and and we had CBS All Access, which became
00:13:31
Paramount Plus. We had our own version of this,
00:13:33
but it was built on, not. Just workout.
00:13:34
I think you're a pretty important guy.
00:13:36
I ran digital for them. So, but they, you know, for
00:13:40
decades they had the upfronts, which, which was, you know, for
00:13:43
CBS, we had Carnegie Hall and, and the whole AG community would
00:13:47
come in and you would, you would tell them the new shows for the
00:13:49
next fall season and all the stars would show up and they'd
00:13:53
have a party afterwards and they would be there to take photos.
00:13:56
I think they called it the petting zoo where you would
00:13:59
have, you would have the logos over every show and all the
00:14:01
stars would be there to take photos with all the press and
00:14:03
all the advertisers that would be in attendance.
00:14:05
The new fronts was a version of that for Hulu and YouTube and
00:14:09
and Netflix, you know eventually once they had advertising and
00:14:11
others. So yahoos version of that is a
00:14:14
little bit different because we do have shows so but they are
00:14:18
very much part of our verticals. So for sports and for finance,
00:14:22
we have nonfiction content that is, you know, really trying to
00:14:26
help you with the decisions you're making, the goals you're
00:14:29
trying to achieve in those categories.
00:14:30
So they're we're, we're not breaking news, we're not getting
00:14:33
scoops. We are providing context in that
00:14:36
category. So we have Kevin O'Connor, who's
00:14:38
the number one NBA podcast. We have Ariel Helwani, who's the
00:14:41
number one MMA podcast, and Brian Sauzi and our our people
00:14:45
on on Yahoo Finance and others. So we do about 60 hours a week
00:14:50
of original programming now for sports, almost that much for
00:14:52
finance. But for the most part, we also
00:14:55
have, you know, this whole network of other products,
00:14:57
right? We're a product company at the
00:14:59
end of the day. And so our new front is really
00:15:02
talking about all the new products that we're launching
00:15:05
with the new Yahoo, all the things that have evolved in our
00:15:08
audience. You guys are everybody's doing.
00:15:10
Them this is just your own individual.
00:15:12
OK, but do they happen? They all happen.
00:15:14
Around different places. You all rent out CBS in the old
00:15:17
days. It's the same period of time.
00:15:19
Everybody's sort of NBC V Radio City.
00:15:21
You would all have. So we have a place a few blocks
00:15:24
away from here in midtown and tomorrow night at six O clock to
00:15:28
8:00 PM we will have two dedicated hours for Yahoo with
00:15:32
me kicking it off. Somehow been convinced to wear a
00:15:36
cowboy outfit on stage to start it off in the big opening and
00:15:41
all our GMs get up and talk about all the all the parts of
00:15:44
the company. Are you optimistic about
00:15:47
advertising in language models in chat bots?
00:15:51
Well, that one's a different. That's a whole new thing.
00:15:53
Well, I just think I'm talking. Like advertisers on the mind
00:15:57
that's, that's just a big question right now so.
00:16:01
Yeah, we can. In way back in the day, my
00:16:03
company, we were the first ones to switch from Overture to
00:16:06
Google Adwords. Like I go way back in in the
00:16:08
search category and I'm very passionate that that search
00:16:14
advertising does need to cross the chasm to AI search in that
00:16:19
while it won't be one for one efficient, meaning, you know,
00:16:23
keyword targeted link blue links that look just like the
00:16:27
editorial links. We're a great business model
00:16:29
for, you know, 25 years now. It will look a little bit
00:16:34
different and probably will be somewhat less efficient in this
00:16:37
new UI. But I think actually the
00:16:39
characteristics of the two are actually more similar than
00:16:42
people would think. And there's a little work that
00:16:44
needs to be done to, to bring them over and how you're going
00:16:46
to match keywords and how you're going to render it.
00:16:48
But I actually think, you know, this new environment will be
00:16:52
more similar to to traditional search than people think in, in
00:16:56
that way. With Yahoo Scout, we actually
00:16:59
one of the core values of the product has been that
00:17:03
forgetting, sorry, forgetting a very Italian with how I'm
00:17:07
talking, forgetting, you know, the, the search advertising part
00:17:13
of it for editorial for, you know, you know, just the
00:17:17
traditional listings themselves for the results themselves.
00:17:21
What we've done with Scout is we've actually put a lot of work
00:17:23
into linking back out to publishers.
00:17:26
So all in our results, which if you go to scout.com, you'll find
00:17:32
our own, you know, Yahoo powered AI search products.
00:17:35
The results are really good. They're really unique and I can
00:17:38
talk about why that is, but also in the UI you'll see a lot of
00:17:41
blue highlights all through. And that's what publishers like
00:17:45
me want, right? We want the old world where it's
00:17:47
like send us traffic. And in some ways you're sort of
00:17:51
saying that aligns with advertisers because they would
00:17:53
like people to actually go to their sites, Whereas the, the
00:17:57
sort of ethos of a chat bot is the answer is here.
00:18:01
And so I guess, do you think, well, you, you maybe disagree
00:18:04
with that, but that is certainly the big, you know, anthropic
00:18:07
open AI, they are trying to solve the problem in the box.
00:18:10
Like, do you think that type of ad will send people elsewhere or
00:18:14
advertisers are going to need to be content with?
00:18:18
Well, they got exposure to me sort of in in this context, They
00:18:21
didn't necessarily go to my website.
00:18:23
We'll start with our editorial results at Scout.
00:18:26
They link elsewhere so our results are by paragraph, but
00:18:31
every now. And then contrary.
00:18:33
To it, OK, but think about why that happened.
00:18:35
So first of all, let's start here.
00:18:37
It's extremely important to link out to the publishers who
00:18:40
created this content in the 1st place.
00:18:42
Without it, whether you're a large language model or anybody
00:18:45
else, you won't have any content to bring together in a answer to
00:18:49
provide value to the user, right?
00:18:51
So those publishers surviving and thriving is extremely
00:18:55
important to everyone involved. Why didn't that happen with the
00:18:59
first generation of chatbot AI search results?
00:19:04
I just don't think they prioritized it.
00:19:06
Everyone who was involved was not a traditional search product
00:19:09
manager or engineer. They all came out of academia,
00:19:12
right? Whether they were deep mind and,
00:19:14
and or then went to open AI, that was their background.
00:19:17
And so the, the UI, not surprisingly, looks like an
00:19:21
academic research paper. And they were, they had sources
00:19:24
at the end that were, you know, footnotes essentially that you
00:19:27
would find in a research paper. And so not surprisingly, they're
00:19:30
not very prominent in the UI. Not that many people click on
00:19:34
them. I mean, I think it's less than
00:19:35
like 1% of, you know, AI search results in a Click to one of
00:19:40
those sources. And I think that was just a, you
00:19:44
know, trick of how the UI of this first generation was done.
00:19:48
Google was caught flat footed, I think by the launch of Chacha
00:19:50
PT. And of course, because they are
00:19:54
by far number one in search, they had to counter that UI that
00:19:58
Chachu is helping define, define the category.
00:20:02
So whether or not they wanted to embrace it or or not, they they
00:20:05
had to. And so that's why Gemini's
00:20:07
first, you know, version of this looks like that with Scout, we
00:20:11
are trying to take it a different direction.
00:20:13
We're much smaller. We have less at stake.
00:20:17
And we I think have have the luxury of being able to employ a
00:20:22
different UI and kind of help show the way on that.
00:20:26
I do think eventually all AI chat bots need to for the their
00:20:30
own good meaning like being able to monetize without subscription
00:20:35
because that's not going to get you all the way.
00:20:36
There are going to need to find a way to link out both because
00:20:41
it's good for the health of the Internet, good for the health of
00:20:43
publishers without whom we won't have this content.
00:20:45
And then, yeah, I do think that there will be paid versions of
00:20:48
that that will look and operate somewhat like traditional search
00:20:52
advertising. And without linking out, you
00:20:54
won't be able to do that either. Right.
00:20:57
I, I never, I hadn't connected the publisher piece to the
00:21:00
advertising piece. That gives me the most hope that
00:21:03
these models need to push out a little bit because they're going
00:21:05
to need to push out for their advertisers.
00:21:07
And so maybe then they push out. So I, I can see that, but I
00:21:12
don't know, the tech industry has gotten so extractive like
00:21:14
this, this view that you're articulating of we need to take
00:21:17
care of the underlying like media organizations or content
00:21:21
providers so that we have the content to aggregate that.
00:21:25
That does seem like sort of the Yahoo Spirit like of yesteryear,
00:21:28
but it feels like that's sort of been forgotten in a world where
00:21:31
Google is just not linking to any.
00:21:33
Like, I don't, I don't know, I don't see the rest of the tech
00:21:35
industry embracing that sort of we need to maintain sort of all
00:21:40
the fuel that we're using underneath.
00:21:42
Well, I think another thing will drive them that way, which is I
00:21:44
think it will wind up being shown over time that it is
00:21:48
better for users as well to have the sources show up more
00:21:53
prominently. In our case, not only do the
00:21:55
blue highlights are they linking out, but if you hover over them,
00:21:57
you see the source. And that's just extremely useful
00:22:01
for someone, you know, doing the research.
00:22:03
As you know, these things hallucinate.
00:22:05
They're not they're, they're dated.
00:22:06
Sometimes they don't always provide the right information
00:22:09
depending on the category and the number one category for
00:22:11
Yahoo Scout so far, you know, only a month and 1/2 in is
00:22:15
health. You do want to validate these
00:22:17
sources and not just take for granted that the engine will is
00:22:21
providing an accurate result. So I actually think it's good
00:22:24
for publishers. I think it's good from a
00:22:26
monetization point of view, but I think most importantly, in a
00:22:29
lot of ways it's better for the user as well just because it
00:22:33
has, it is still early. We're still only three years
00:22:35
into this, and that UI that has become the default I, I, you
00:22:40
know, was a, you know, a fluke of how it was launched.
00:22:44
I mean, I don't even think they anticipate how popular it was
00:22:46
going to be if you go back and do this.
00:22:48
Yeah, you know. Everybody, yeah, I'll ChatGPT.
00:22:50
I was like Oh my God. Wow, this is crazy.
00:22:52
That's right. And so, so that's, that isn't
00:22:55
the UI that had to, you know, that had to be.
00:22:58
And I don't think it's what Google would have launched if
00:23:00
they had come out the gates first.
00:23:02
Do you have a view on how copyrights going to net out?
00:23:05
I mean, the Trump administration just put out their sort of AI
00:23:09
guidelines, which I thought were somewhat thin.
00:23:11
I don't have a full view on all of them.
00:23:13
But one of them was sort of like, we should let the courts,
00:23:16
I guess, decide on copyright. Like, we're not going to come
00:23:18
down super hard. And that that's going to play
00:23:20
out if it's into what we were talking about, just because if
00:23:23
you're not, if you're not emphasizing pushing people to
00:23:27
links, it feels more and more like you're just stealing their
00:23:30
IP if you're just using it for your own answer.
00:23:32
Like, have you done an analysis on how much you think sort of
00:23:36
models could get hit with copyright at the end of this or
00:23:39
they're just too ubiquitous and they've just won by being like,
00:23:42
so dominant? Yeah, I'm not.
00:23:44
Yeah, not a lawyer. Just play one on TVI.
00:23:47
Have lawyers running around. Well, it's interesting.
00:23:49
I mean, I do think that for 30 years the social contract of the
00:23:56
Internet and of search was that a search engine would crawl you
00:24:00
unless you with robots TXT you decide you didn't want to be
00:24:03
crawled for whatever reason. And then search engines
00:24:05
respected that or were supposed to.
00:24:08
And that the social contract was that they would take a snippet
00:24:11
and the link and they would link back out to you.
00:24:14
And then if you wanted to have other information like, you
00:24:17
know, lyrics, you would actually or, or images, you would acquire
00:24:20
those from a structured data provider and then you would add
00:24:23
those to the to the page. So, but the reason you wanted to
00:24:26
be crawled and you want it was you want traffic.
00:24:28
And that was how you had a business so that you could keep
00:24:31
creating that content. So if you start to go to a world
00:24:33
where that content is is taken, digested and interestingly, not
00:24:39
even for the answers, but to be able to predict it's auto
00:24:43
complete, you know, to predict the next word or a letter or
00:24:47
token. I mean, it's just, you know,
00:24:49
that was never contemplated when you were coming up with this.
00:24:51
And I feel for the companies who are really based on Evergreen
00:24:56
content. You know, with Yahoo, we are one
00:24:58
of the top 20 publishers whose, you know, content shows up in an
00:25:03
LLM. So we are on weirdly on both
00:25:05
sides of this. With deals you struck or just.
00:25:08
No, no, no. They're just taking, yeah.
00:25:11
We are not suing over it. No, we haven't.
00:25:13
We have not. We've not done that at this
00:25:16
point, but. You know when the you have a
00:25:18
clock, you're like, oh, we have to decide by.
00:25:21
No, I mean it's, it's, you know, it's interesting that that even
00:25:25
the people who are being paid don't believe that they're going
00:25:28
to be able to make enough of a living that way.
00:25:31
So that's the other interesting thing happening here.
00:25:33
I think only 20% saw a survey. Only 20% of publishers believe
00:25:37
that they can they can make enough money by selling their
00:25:41
content to large language models.
00:25:43
Well, like. The publish some publishers
00:25:46
struck what seemed like a huge deal with Anthropic.
00:25:48
I don't know the numbers, but then if you look on a per book
00:25:50
basis, it's it's not like you could write books for a living.
00:25:53
Just being like, Oh well, this book is going to well.
00:25:54
That's books. I mean just talk about web
00:25:56
publishers. I mean, 10s of 1 is
00:25:58
probably not going to cut it for the, the level of businesses
00:26:00
that, that these companies are running.
00:26:03
So something else has to happen. I, I believe what needs to
00:26:06
happen is, is they need to, before it's too late, do the
00:26:10
right thing and link out and do the, let's do the right thing
00:26:13
for the users and do the right thing for publishers.
00:26:15
So let's start with trying to get there first, you know,
00:26:17
before we worry about, you know, just just throwing in the towel
00:26:20
and, and saying this is the way it's always going to be.
00:26:23
But I, I really, you know, again, it's not just Pollyanna
00:26:26
like, Hey, let's do this and try to be good, the good guys here.
00:26:30
We are trying to be the good guys on that for sure.
00:26:34
We think it's the right thing to do.
00:26:35
But I also really do find it really valuable in the UI.
00:26:37
And I think if you use Scout you'll feel the same way.
00:26:41
Finance is obviously a huge important vertical one that I
00:26:44
think about a lot. Do you, how much do you think a
00:26:50
retail trader is going to become smarter because of LMS or do
00:26:53
you, how much are you trying? Are you doing anything to give
00:26:55
them advice based on what language models think and like
00:26:58
what do you see in terms of, yeah, just like the retail
00:27:01
trader and what they want or financial environment in this
00:27:04
new world? Well, one dirty secret of that
00:27:07
is is that, you know, of course, Yahoo Finance for 30 years, you
00:27:12
know, people have tried to copy it, tried to kill it.
00:27:15
Google themselves. Interestingly, the person who
00:27:17
started it for Google is on my board at Yahoo now.
00:27:21
Katie Stanton. But but it's interesting how
00:27:25
resilient that property has been over the years.
00:27:27
And that's before, you know, I think we've really tried to make
00:27:29
it better and have some other things coming that I think would
00:27:33
be a lot better. Again, you fall into the the an
00:27:38
area where the most basic features are the ones that are
00:27:43
used the most. And so that deep, deep research
00:27:45
is used by a smaller and smaller percentage of the charter base.
00:27:48
We are, yeah. But but really, you know, stock
00:27:52
quotes and news wound up being at the very top of that pyramid.
00:27:56
And we are building things, you know, using AI, using Scout that
00:28:01
allow, you know, people to go deeper and do hardcore research.
00:28:05
And, and we do have things in the pipeline for that.
00:28:09
We've also as part of the launch of Scout, we, we did launch
00:28:12
features straight into every one of our products so far, not just
00:28:15
finance, but fancy Sports News, e-mail, everything has got some
00:28:20
bit of Scout, we call it the Scout intelligence platform kind
00:28:23
of pushing features into it. And one is, you know, explain to
00:28:27
you like what changed about the stock, why, you know, why is it
00:28:30
trending in this direction, things like that.
00:28:33
But yes, I I think, I think there is a future of much richer
00:28:38
detailed products in every category that AI can help drive.
00:28:43
One we've, we've sort of hit on twice the situation where the
00:28:48
consumer doesn't want sort of the Super cheap features or like
00:28:51
they like advanced search. And then on finance we're
00:28:53
talking about, you know, it's like they, they want a couple
00:28:56
things they're not necessarily doing like intense hardcore
00:28:59
research. I feel like in sort of the AI
00:29:02
consumer, somebody I think had a tweet I'm stealing, I'll have to
00:29:06
go find it and throw it up on the screen.
00:29:08
But basically the idea is that I think there are a lot of people
00:29:10
who think consumers want AI to like save them time like like in
00:29:16
business or like make me sort of more efficient, better at like
00:29:19
calendaring or something. But it, but in a lot of cases
00:29:21
what they actually want is sort of to waste time sort of more
00:29:24
enjoyably. And they want AI to be like fun
00:29:27
and and lively. And I, I guess I'm just getting
00:29:30
at like, what do you think the consumer really wants from AI?
00:29:34
Like where, where do you think there are real needs that LLMS
00:29:38
can satisfy versus right now they're sort of just getting
00:29:41
thrown? Yeah, things that researchers
00:29:44
built sort of natively to them and stuff that's working for
00:29:47
enterprises but isn't, doesn't really have like the consumer.
00:29:50
It it is, it always has been category by category.
00:29:54
So I think AI now is, you know, horizontal across everything
00:29:58
that's happening out there. And it depends.
00:30:00
So, so yeah. And, and I don't know what part
00:30:05
of it you're using for entertainment is that like it's,
00:30:07
it's your buddy and you're having you're having
00:30:08
conversations. With me, I had the sort of like,
00:30:11
oh, we're going to create like slop videos.
00:30:13
My my wife has recently been sending me these.
00:30:17
I don't know how much TikTok you waste your life watching, but
00:30:21
they're a very viral TikTok at the moment is fruits like, you
00:30:25
know, arguing with each other, like cheating on each other.
00:30:27
It's like as if it's like the bachelor, but it is fruit and it
00:30:31
is like it's the first time I I don't I found it terrible.
00:30:34
I was like, this is I don't know why you're watching this, but
00:30:36
it's like slop. You know, he's sort of hitting
00:30:39
story in a way. Like I feel like this sort of
00:30:41
like time wasting video is where it like can be sort of tested
00:30:46
enough and it's just like throw shit at the wall and see what
00:30:48
resonates with people. I don't know, I.
00:30:50
That makes sense. That's that's just incremental.
00:30:52
I mean, Amazon just announced AI Studios.
00:30:54
I think Albert Chang was formerly ABC and Disney and is
00:30:59
now running it. So there obviously is a future
00:31:03
for that. It makes more sense in short
00:31:04
form videos, right? Lower consequences and and it's
00:31:08
more time. Burning spin up the whole movie.
00:31:10
Yeah, but Tik Tok's not new or, or Reels is not new to how you
00:31:13
or YouTube Shorts to how you've been spending time, you know,
00:31:16
that's consumed much, much more of your entertainment time than
00:31:19
it used to. So, you know, I think that makes
00:31:21
sense. We we tend to operate much more
00:31:23
in categories where they you're trying to get things done.
00:31:29
So you do have a little bit more of that professional?
00:31:31
Well, again, it's some of them are utilities.
00:31:34
So I would say e-mail search, maybe weather are in the
00:31:38
utilities category. We do have content category,
00:31:41
sports and finance. But again, in finance you're
00:31:43
trying to make more money or save money.
00:31:46
Sports is the one where, you know, obviously that's
00:31:48
entertainment, but Yahoo users are using it to play fantasy,
00:31:53
beat the snot out of their friends.
00:31:55
Take it too. Seriously, that win money, you
00:31:57
know, they're, they're actually using it for, for these things.
00:31:59
So not to be cheesy about it, but we did have to reset the
00:32:03
mission of the company. When I got here, Yahoo was the
00:32:06
turn around. It had been owned by Verizon for
00:32:08
many years before that, serving whatever purposes they bought it
00:32:11
for. Before that, it was a struggling
00:32:13
public company having to make really hard choices about what
00:32:17
it was going to do to survive and, and try to grow against
00:32:20
some really hard headwinds of Google and Facebook and YouTube
00:32:22
and, you know, social media, other other things.
00:32:25
So, you know, we had a, the benefit almost of being able to
00:32:31
only focus on what was good about Yahoo.
00:32:33
We didn't have to worry about the rest of it.
00:32:34
Why is Yahoo Finance still #1? Why is Yahoo Sports top two?
00:32:38
Why do so many people want aggregated news?
00:32:40
And like, what can we do to lean into that?
00:32:42
Why is Yahoo Mail so pop still so popular?
00:32:44
And it it the spine of that book is is always.
00:32:47
What is it? What's the?
00:32:48
Answer it always goes back to Jerry and David's original
00:32:51
founding of the company. They were trying to be the guide
00:32:53
to the Internet. It says it right on the original
00:32:56
page, Jerry and David's Guide to the World Wide Web.
00:32:58
And that started because Andreessen had started Mosaic
00:33:01
and it was originally for, it said on the page for Mosaic.
00:33:06
And it was because that you had access to the Internet, but you
00:33:08
didn't know how to find anything.
00:33:10
And then that evolved into doing things.
00:33:12
And Yahoo is your home base, you know, for that.
00:33:14
So we, you know, did not have infinite resources taking this
00:33:18
over. Our job to do is to make sure we
00:33:20
understood that people were using Yahoo for, and that was
00:33:24
for trying to accomplish these goals in whatever category it
00:33:27
was. So all of our focus and road map
00:33:29
started to go behind that and that's when we started move out
00:33:32
things that didn't. Right, but it's still rooted in
00:33:34
the Internet or it's. All, all Internet, but you're
00:33:37
using the Internet to accomplish those things.
00:33:40
So when it comes to AI, we just launched literally today as an
00:33:45
example in e-mail and you talk about entertainment versus
00:33:48
calendar and all that. But we launched a product called
00:33:50
Planner and it does go in using Scout technology and extracts
00:33:55
all these like next step things that are demand need to demand
00:34:00
your attention, that you need to know, you know, a deadline,
00:34:03
something for your kids school, a flight doctor's appointment.
00:34:07
It extracts that out and automatically adds it to your
00:34:09
calendar. If you're, you know, or to A to
00:34:12
do list and, and Yahoo and you're, you know, we have that
00:34:15
feature as well. And that's using AI to remove
00:34:19
steps so that you can get things done better and faster that you
00:34:22
didn't have to think about. Like, I think that's a good one.
00:34:24
I had my first fantasy baseball draft of the season last night
00:34:28
and if you are a Yahoo Plus subscriber, it will be your
00:34:33
research assistant and help you do a better job in your draft.
00:34:36
So. So it really is category by
00:34:38
category. I I buy that Yahoo is more of a
00:34:42
friend to the Internet and some of these LMS because it's like,
00:34:46
OK, we want to get people to the good stuff out there so they can
00:34:50
find it. And that's really what the
00:34:51
business has been about. But has that raw material just
00:34:55
gotten, like, worse over time? Like, is the Internet just
00:34:58
degrading? It feels like media is getting
00:35:02
worse. You know, I think of like, even
00:35:05
Reddit, which for a while I think there was a meme, it's
00:35:07
like, oh, if you wanted a good Google search result, you just
00:35:09
search what are they saying on Reddit?
00:35:11
Which I think is somewhat true, but to some degree I worry more
00:35:15
and more that Reddit posts are becoming like was this an LLMAMA
00:35:19
or or is this a real 1? I don't know.
00:35:22
Do you worry about just like the raw materials that you rely on
00:35:26
on the Internet getting worse right now?
00:35:29
I mean, I hate to always bring it back to this isn't new, but
00:35:32
you know, if you go back to my first time and search Danny
00:35:36
Sullivan, if you remember Danny, but he had the biggest search
00:35:38
engine conference 5000 people would go and it those of us you
00:35:43
know who ran search engines, we'd be on stage and and and
00:35:47
talking, but in the audience were all the spammers, all the.
00:35:51
The whole SEO, how can we take advantage?
00:35:52
Of the whole SEO community and if you went outside or or to the
00:35:55
bar afterwards, everyone's together.
00:35:57
And that game of trying to send traffic for their own purposes,
00:36:02
that's not new. And yeah, there were there are
00:36:05
really good SE OS who are trying to be helpful for publishers and
00:36:07
there's ones who are trying to gain the engines to, you know,
00:36:10
drive it to to, you know, lower tier traffic.
00:36:14
That is not a new game. I don't think AI slop.
00:36:17
I think that's another part of it.
00:36:18
Think about the decade before clickbait.
00:36:20
I mean, it's just, you know, all these attention vampires out
00:36:24
there, like trying to think about how headlines change over
00:36:26
the years and the ones that you would have written 10:15 no,
00:36:29
probably 1520 years ago to over the last 10 years after
00:36:33
BuzzFeed. And like how every, even New
00:36:35
York Times headlines became very like BuzzFeed type headlines.
00:36:38
So this trend has been out there.
00:36:40
But I would say there's more good content than there's ever
00:36:42
been certainly than 20 years ago on the Internet just by sheer
00:36:46
volume. If you think about not only is
00:36:49
the content on the Times and Washington Post and CNN and BBC
00:36:53
and you know, all these places, like there's amazing content in
00:36:57
these places, but then with Substack and then with YouTube,
00:37:00
I mean, it's, it is endless how much great content is out there.
00:37:03
So I, I think it is our job to help you find the good stuff and
00:37:07
to, to be your guide through that digital wilderness and
00:37:10
then, you know, avoid falling into, you know, off a Cliff or
00:37:13
stepping on a rattlesnake or whatever.
00:37:14
We have to help you do to, to get through it all that that is
00:37:17
historically our job. And we have to, you know, do a
00:37:20
good job of that. That job will get harder over
00:37:23
the next 30 years, but it's not a new job.
00:37:26
I think I agree with you that overall it's gotten better.
00:37:28
I mean, YouTube explainers that whole category is sort of like
00:37:32
the, you can go to YouTube for sort of any question you have
00:37:34
about the world. I think the New York Times has
00:37:38
gotten better, but obviously, like local papers have generally
00:37:41
gotten worse. It's hard.
00:37:43
It's hard to score overall. I mean, I yeah, there's some.
00:37:47
There's some puts and takes. In this generation, unless like
00:37:50
Jim Prosser, APR Guy in Silicon Valley, who I respect has a
00:37:53
blog, he he basically argued that like the SEO optimization
00:37:58
for LMS category, I forget what people are calling it, but yeah,
00:38:02
but he said he thought that was sort of fake or like that, that
00:38:05
there isn't as much to be done there.
00:38:08
Do you agree with that? Honestly, I, I'd follow a lot of
00:38:11
SEOSII, you know, came out of that industry.
00:38:15
I, I, I know a lot of those people.
00:38:17
I read on Twitter all day long, everything.
00:38:19
They're just trying to do what they can to adapt and to get the
00:38:25
that that traffic was pulled out from under them in a lot of
00:38:28
extremely important categories that that have been.
00:38:31
And and to your point, you know, not really with any heads up,
00:38:38
you know, this was just done in the dark of night and all of a
00:38:41
sudden it's like here are these, here are these.
00:38:42
Products that everyone choose on the SC, so they're saying is
00:38:45
there money to be made helping people optimize for language
00:38:48
models right now or you think that's not going to be a big
00:38:50
business? I mean, if it is the traditional
00:38:54
large language model with, with footnotes at the end, it's going
00:38:57
to be a hard, a hard one. But but yeah, it's the the
00:39:02
industry will have to adapt to prioritizing traffic or it'll be
00:39:05
hard. I mean, I do watch those
00:39:07
discussions that they're having about trying to figure out like,
00:39:09
oh, if you, you know, 'cause what they're hearing from Google
00:39:13
and others is, you know, it's the same, do it the same.
00:39:17
You know, you have to optimize for these characteristics in
00:39:21
your on quality and speed and freshness and all these things.
00:39:26
But at the end of the day, if the UI does not enable someone
00:39:29
to easily click out or click out at all, it's going to be hard,
00:39:34
right? Hard to see value from it's.
00:39:37
Not hard to get traffic, Yeah. And by the way, I also think
00:39:40
users have been trained. But, and this is where I go to,
00:39:43
we're only three years into the next 30 years.
00:39:45
So I think it can be evolved over time now, but it's not
00:39:49
going to happen overnight. They have been trained initially
00:39:52
with these products to not click, you know.
00:39:56
Definitely the my biggest concern rather than like spam
00:40:00
for language monolayer. Well, not my biggest concern.
00:40:02
A concern is just whether the consumer is savvy enough, right?
00:40:06
I think on Twitter especially, you see all this like AI writing
00:40:12
that goes so viral and it just feels like.
00:40:16
Who is the speaker? Like, the markets are moving,
00:40:19
you know, off of reports that feel like they could have been
00:40:22
written by AI. Yeah, although the mission for
00:40:25
for Twitter is Wild West, it's free speech, anything goes.
00:40:30
We're not going to curate it. And that's a good thing.
00:40:33
So that's probably not the right place to go if you want to, but
00:40:35
that's. What?
00:40:36
I would argue those, yeah, but I would argue.
00:40:38
Those users are getting what they're they're asking for in,
00:40:40
in using that product. I mean, I use it, you know, all
00:40:43
the time and you just have to know what you're getting
00:40:45
yourself into. It is, you know, they, they have
00:40:47
been launching more features to, to maybe help you curate it
00:40:50
more. And but you know, look, every
00:40:53
one of these product feed based products has is always trying to
00:40:56
get you to the 4U column. You know, they're, they're all
00:41:00
about that and that's the one that's most susceptible to that.
00:41:03
What I mean your, but your answer for this problem is to
00:41:07
trust sort of news outlets, right, Which you then sort of
00:41:10
drive traffic like you're not personally going and saying like
00:41:13
who's real, who's not, what's true, what's not your.
00:41:16
And we definitely are when it comes to yahoo.com or Yahoo
00:41:20
News. So what you're saying which
00:41:22
outlets are true, right or like you?
00:41:25
You we are trying to be purple in terms of left and right.
00:41:28
And I would always remind people there's only one category.
00:41:31
When people say they don't trust the media, they're not talking
00:41:33
about sports or food critics or movie critics or, you know,
00:41:37
people who cover the real estate.
00:41:39
It's politics is the is the dividing category and, and, and
00:41:44
so for that one, we really try to especially for people who are
00:41:47
not logged in, just starting out, not move them one way or
00:41:49
the other. Now you will by the algorithm be
00:41:51
moved one way or the other based on your preferences over time.
00:41:54
I think that's totally fair as long as the, the source is a
00:41:57
trustworthy source. We, you know, we are partnered
00:42:00
with hundreds into the thousands of publishers.
00:42:03
So we we have vetted these sources that you will see
00:42:06
showing. Up right you're deciding is
00:42:07
Reuters. Trustworthy.
00:42:08
The only thing that shows up on Yahoo is someone we've done a
00:42:10
partnership with when it comes to, to news, when it comes to
00:42:14
the to Yahoo scout or AI answer engine, you know, there's,
00:42:18
there's all kinds of data that's coming into that, including
00:42:21
vertical publisher data. We also grounded on on Bing and
00:42:27
you know, we're, we're, you know, it's, it's based on usage
00:42:30
data, query data, you know, all, all the things that Yahoo
00:42:34
uniquely has over time. It's one reason our answers are,
00:42:36
are pretty unique. But in that way, it is curated.
00:42:39
It is not just a random pull. By digesting the entire Internet
00:42:43
and then predicting the next, you're striking partnerships.
00:42:45
Yeah, we're actually trying to have a high quality in our
00:42:49
answers. I mean, the news business can
00:42:50
feel essential, a trend we haven't touched on as much, you
00:42:53
know, like prediction markets. I think there was a reporter who
00:42:56
was worried he was getting threatened over saying that like
00:42:59
a missile had hit something because a prediction market
00:43:02
hungover. And there was an angry reader
00:43:04
trying to get him to adjust the article to, I guess, help the
00:43:09
bet. So, you know, prediction markets
00:43:11
resolve to some of these news outlets obviously.
00:43:13
Right, it's. Amazing language models.
00:43:16
You know, it's like a lot of materials news.
00:43:19
I mean you you own TechCrunch, which obviously.
00:43:21
No, no, we sold. Yeah, you owned past tense,
00:43:23
that's what I'm saying. Yes, past tense and decided not
00:43:26
to continue owning. Like how much is your view that
00:43:30
yeah, the news business is fucked?
00:43:33
Well, that wasn't it. It was that we were kind of had
00:43:35
one foot in hot water, one in cold water.
00:43:37
Like who are we? We have to choose.
00:43:39
We were bought by private equity.
00:43:40
We're starting back out. Like you just have to focus, you
00:43:43
know. And so for us over time, it was
00:43:47
that we're not in the breaking news business.
00:43:49
So we aren't having reporters on the front lines, you know,
00:43:54
reporting the news out from there that we will aggregate
00:43:56
that news and send them traffic. And we're glad they do it.
00:44:00
But that's not that's all we do. We are our our content is for
00:44:04
context to the things that people breaking the news, you
00:44:07
know, report. And historically, Yahoo Sports
00:44:09
is a good example. The best breaking news reporters
00:44:13
in in basketball, you know, originally Woj, then Shawn's.
00:44:17
They started at Yahoo. You were the breeding.
00:44:19
Ground, it was the breeding ground for that.
00:44:20
But again, within, and I saw this at CBS as well, it's
00:44:22
because we ran CBSSports.com. We had all kinds of, you know,
00:44:27
writers on the front lines and you know, SportsCenter would
00:44:30
barely mention us, you know, in breaking the news and it was in
00:44:34
within 5 seconds, you know, it's already out there.
00:44:36
So we just felt like that's not where we can have the most value
00:44:39
and maybe we don't find the most value ourselves.
00:44:41
Which is it? Like and also why?
00:44:43
Be the but is there value but? Also, why be the 200th player in
00:44:47
that? It's, it's not where Yahoo
00:44:49
should play. You know, we, we, we, we were
00:44:51
literally competing against The Associated Press for news story
00:44:54
that that's not our job. Our job is to send The
00:44:56
Associated Press traffic. Yeah, I mean, you define your
00:44:59
job. I mean, I guess to me the world
00:45:02
is short on breaking news. I'm not saying you need to.
00:45:04
Anybody is obligated to be in any particular business, but I
00:45:08
want to disentangle like do you think breaking news is not like
00:45:11
a great. Business again, I would say we
00:45:16
have to go back to sometimes when people talk about someone
00:45:18
buying our company and they're like, oh, but what do you get
00:45:20
get by regulators. I'm like, I love that we're at
00:45:23
this point in time 31 years that after all Yahoo went through and
00:45:27
how much people in the industry kind of dragged the brand
00:45:30
through the mud that they would think that we would be the
00:45:32
source of, of being, you know, of a regulatory thing with, with
00:45:36
an acquisition that doesn't make sense.
00:45:37
But I would say the same thing here that Yahoo of of all you
00:45:41
know, publishers would be the one that.
00:45:43
That is being why hold us to account on.
00:45:45
That no is is is the one that you're saying is missing from
00:45:48
the front lines of of journalism.
00:45:50
I I think there's a. Lot You're doing far more
00:45:51
journalism than every other tech platform in the world.
00:45:54
Right. We, we do a lot of journalism,
00:45:55
but again, we do it from a certain point of view of like
00:45:58
that is not the breaking news side that I think a lot of
00:46:00
people, I'm not trying to put their.
00:46:02
Resources into. In some ways, it's because
00:46:03
you're the closest to it. You have to.
00:46:05
Bring a turn around is is people, as you know, are saying
00:46:07
like what's the formula for that?
00:46:08
There's not always one people is the most important part, people
00:46:12
who know what they're doing. And then they naturally put a
00:46:14
great strategy in place and build great products and, and
00:46:19
you know, and all that, but, but you know, one part of that is,
00:46:23
is strategic focus. You can't do everything,
00:46:26
especially when you're trying to fix the PNL, you know, build,
00:46:30
get a solid foundation going for the for the future.
00:46:33
That is where we really were coming from and all that.
00:46:36
Would breaking news, I mean it becomes a headache or like you
00:46:39
where are things getting like escalated up to you or just like
00:46:41
the business of breaking news is hard to manage or.
00:46:45
Not for very well. We we made these decisions
00:46:46
fairly always like after we took over.
00:46:49
But. But no, I wouldn't.
00:46:50
We certainly no decisions would make their way up to me.
00:46:54
Yeah. I mean, you remember back at
00:46:55
CBS, I had all kinds of of news brands.
00:46:58
This was my first rodeo running, running a news brand like we
00:47:00
did. We did not get involved.
00:47:01
No. Well, that's why I'm asking for
00:47:03
your sort of overall perspective on news right now.
00:47:07
Yeah, I mean, it's, it's an interesting spot.
00:47:10
I mean, I think the corporate ownership of all these things,
00:47:16
the the, you know, what's being pushed on journalists from an
00:47:21
editorial perspective. And you know, at the same time,
00:47:25
there are we, we work with hundreds, thousands of sources.
00:47:29
You know, I think the ones or the flashpoints are, are always
00:47:32
interesting. There's nothing more interesting
00:47:34
to journalists than other journalists.
00:47:35
And I've always found that too. And, and so, you know, it's it
00:47:41
that is not happening across the board.
00:47:43
There are, there are still the vast majority of sources are,
00:47:46
are not falling into that camp. So I, what I would say is news
00:47:51
sources are definitely seeing less traffic, you know, because
00:47:54
of the way search has been evolving.
00:47:56
So that part of it has been difficult for the industry.
00:48:00
And you know, staffing in a premium way requires premium
00:48:04
CPMS, right? And I think what happened over
00:48:07
the previous 15 years was programmatic CPMS against
00:48:11
premium staffing equals very difficult business model, right?
00:48:15
So now other things have evolved like podcasting and other ways
00:48:18
to, you know, to, to kind of to go with if, if, if you're those
00:48:23
kinds of brands. But yeah, it is it has been a it
00:48:26
has been a hard business and harder for some than others.
00:48:30
You know, New York Times is still doing really well.
00:48:32
There are there are others doing well.
00:48:34
Anyone who's subscription based, and this could be, you know,
00:48:38
really anybody, you have to start with a broad funnel
00:48:41
usually of free. If you're Netflix and you're you
00:48:44
define the category cha chi PT in some ways with that where it
00:48:47
was a subscription model to start with.
00:48:49
The only way you could use it was in many ways was
00:48:52
subscription based. But at the end of the day, only
00:48:57
5% of chats CPT users are I believe are paying subscribers.
00:49:01
So again, back to that broad funnel.
00:49:03
And I think that requires brand that requires scale.
00:49:09
So, so it's, it is, you know, for the few who are able to have
00:49:13
that kind of a a model. And then on the other side of
00:49:15
that, you will have more niche things.
00:49:17
And that is where, you know, historically trade publications
00:49:20
would fit in. They would charge a lot of money
00:49:22
and not have a huge subscriber base, but they're prestigious in
00:49:25
their industry. And I think I'm.
00:49:27
Happy to take on that. Yep.
00:49:29
Mantle and I think that Substack has and and others have come
00:49:32
along in that niche and further that so.
00:49:36
What Yahoo's obviously privately owned Like what?
00:49:39
What can you say about the Business Today?
00:49:42
That's good being private, I think.
00:49:45
Do you want? To brag I feel like.
00:49:47
No, not yet. At some point it's, you know,
00:49:51
look, this was the kind of the white whale of turnarounds.
00:49:55
It was one that, you know, I don't think a lot of people
00:49:57
could thought could be pulled off and it was even, you know, I
00:50:00
feel like we have done that. But I think it's been a 4 1/2
00:50:03
year now journey to get there. And you realize like selling
00:50:06
AOL, like how much have you do the math I guess for sort of the
00:50:11
Apollo deal to to now? Yeah.
00:50:13
So, yeah, I mean what they have said publicly is they've already
00:50:17
returned the investment, the cash investment on it.
00:50:20
So we're we're kind of playing for upside, you know on the on
00:50:23
the rest of it. If you look at it as APE deal,
00:50:25
we're a little, we're very unusual for that, right?
00:50:28
Because we are part of the beast of Silicon Valley, you know, who
00:50:31
we compete with for users, who we sell to, you know, all of
00:50:35
that. It's not a, it's not really, it
00:50:37
doesn't fit the model of a traditional, you know, PE deal
00:50:40
in that way. And then AI, you know, when we
00:50:43
bought the company, it was the middle of like stonks and, and
00:50:46
all that. And then and then AI wasn't even
00:50:49
on the map. And that happened November of
00:50:50
22. We, we spun the company out in
00:50:52
September of 21. So that's changed the whole
00:50:54
game. So yeah, with deals like AOL,
00:50:57
we, we actually sold the content delivery network that they own
00:50:59
when we first joined, we sold the branding rights to Yahoo
00:51:03
back to Yahoo Japan. We first started.
00:51:05
So they're like, and you know, then smaller ones like
00:51:07
TechCrunch and Rivals and Gadget.
00:51:10
So you take those all together and it's been, you know, we've
00:51:12
returned, you know, quite a bit. And then the amazing thing is
00:51:16
that the core Yahoo business, we've really optimized that.
00:51:20
We've gotten the cost in a really good place.
00:51:22
The quality of the people is incredible now and, and they're
00:51:26
all, you know, off and running traffic is growing at a really
00:51:29
good rate. Revenue is growing at a really
00:51:31
good rate, very profitable. But you know, this was a company
00:51:35
that went public in April of 1996, you know, started running
00:51:40
into some headwinds in the 2000s.
00:51:41
And I think for the next 15 years, you know, it was kind of
00:51:44
rough to be be Yahoo, the public company.
00:51:47
If you think about the, the famous dealer, they didn't sell
00:51:50
for 44 billion to Microsoft. That was 2008, you know, and it
00:51:54
went on another almost 10 years before selling to Verizon.
00:51:58
So that was a long time. So there it is good to be.
00:52:01
I've worked, I've done 2 startups for it that I founded
00:52:04
or Co founded. But you know, the majority of my
00:52:07
careers want to be in public companies or or large companies.
00:52:11
I'll just say there's some benefit to being public.
00:52:13
I mean to being private to do the, you know, to do all the
00:52:17
things to fix it because once you go public as you know like
00:52:20
it is all about, you know, you, you are held to account quarter
00:52:23
by quarter and you better understand what those next, you
00:52:26
know five years are going to look like before you get there.
00:52:28
What? And it's also hard to make
00:52:30
changes in public. What is what is a win look like
00:52:34
for Scout and sort of your AI strategy?
00:52:37
How much is it about showing the industry what it should be doing
00:52:41
and then playing with the industry versus you want this
00:52:44
just be sort of a huge product on it's own?
00:52:46
Like what? What is the goal with your AI
00:52:48
strategy? So if you go back to the turn
00:52:51
around strategy, it was a couple years of transformation work to
00:52:55
the org to the team. And then once you got through
00:52:59
all into the revenue model and the the revenue engine, you
00:53:02
know, we we migrated from the old revenue engine to to a new
00:53:06
one that that was just you know was better.
00:53:09
And then once you get through that, we are in the right OK to
00:53:12
then go and fix all the products.
00:53:13
So then we went through that phase and now we're in the shots
00:53:18
on goal period. We've actually we have the
00:53:20
resources like it's working we're.
00:53:22
Making money. We got rid of the stuff that
00:53:23
doesn't. Make sense, do something with
00:53:24
this audience and that is, you know, very loyal to us and the
00:53:28
places that are very logical not to go do some rando innovation.
00:53:32
Something has nothing to do with the spine of the book for for
00:53:35
Yahoo, but things that our users would find unique, uniquely
00:53:37
valuable, but also your asset base.
00:53:40
So I would, I would say there's two big ones right now.
00:53:43
One is Scout. The others are DSP, our demand
00:53:46
demand side platform in the ad tech space, which is growing in
00:53:51
a really good clip that competes with people like trade desk and
00:53:54
DV 360 from Google and Amazon and a lot of smaller players.
00:53:59
We're the largest private DSP and that is essentially, you
00:54:03
know, you can target talking about the new fronts, you can
00:54:06
target on Yahoo. Oh, no, we will sell ads to you
00:54:09
directly to to go on to our properties.
00:54:12
You can also using, you know, our data then go through us to,
00:54:18
you know, have ads on other people's properties.
00:54:21
And so that that has been a really good business.
00:54:23
We shut down to ad tech businesses to focus on that one.
00:54:26
And we think that could be potentially worth, you know,
00:54:29
billions of dollars if we get that right.
00:54:30
So we're in business. Is it I mean, it's a lot of top
00:54:33
line revenue. You do, you know, there there's,
00:54:36
there's a lot of revelation. That goes back, right?
00:54:40
Yeah. So so you you do and how much
00:54:43
are people using out of traffic that?
00:54:45
Totally advertise off of Yahoo or it's mostly for people who
00:54:47
are already doing a lot of Yahoo.
00:54:49
Ad no, it's for only 9% of the impressions in that network are
00:54:54
are Yahoo. So that's also our competitors
00:54:57
will say you're only buying Yahoo and try to throw us under
00:54:59
the bus that way. It's actually 90% or off of
00:55:02
Yahoo and so Netflix and Spotify and many many.
00:55:05
That's the DSP. Work with the Yahoo DSP Scout is
00:55:09
interesting because there's a lot of great reasons to do it,
00:55:12
which is why we built it ourselves.
00:55:14
So Yahoo had outsourced search since 2009 to Bing and same same
00:55:20
as Yahoo did originally. So it's almost like an
00:55:22
enterprise search deal. And Bing did that to build up
00:55:26
the volume in their network, right?
00:55:27
The way search advertising works is it's not about the technology
00:55:30
as as much as it's about volume. The NTH advertiser will, you
00:55:35
know, increase prices in the auction, they'll expand
00:55:39
coverage, they'll increase relevance, which will help the
00:55:41
click through rate. That is really the secret is the
00:55:43
NTH advertiser goes to Google. So if you put Bing and Yahoo
00:55:47
together, you know, for the Bing marketplace, it actually
00:55:50
increases the likelihood the NTH advertiser will also come into
00:55:53
the into Bing. So that's why Yahoo and Bing
00:55:57
have been partnered for so long. But when it came to this, we
00:56:01
looked at the category about a year ago because we're coming
00:56:03
through that transformation period felt like we, you know,
00:56:06
this is the next obvious thing to focus on given search such a
00:56:09
big part of our portfolio still, but outsourced it.
00:56:13
And we looked at everything. We looked at every outsourced
00:56:16
possibility that you could think of and it came back to the idea
00:56:19
that given all the data, 30 years of search history, all of
00:56:23
our usage data, our user data, the publisher data, you know,
00:56:27
all of it, we actually were sitting on this goldmine of of
00:56:31
data that you would you could use to create a really good AI
00:56:34
answer engine. Right.
00:56:36
You know. What I mean partner, partner
00:56:37
with an index because you have to ground it for accuracy.
00:56:40
But if you had that and and then it became very obvious that, you
00:56:45
know, we already had a big search team that really weren't
00:56:48
building index search that we could repurpose for this.
00:56:51
So we we tried to figure that out.
00:56:53
And then last summer we bought the company, a company, a start
00:56:56
up run by a guy named Eric Feng, who's kind of a luminary in the
00:57:01
industry. He he, you know, back in the
00:57:03
day, he worked at Microsoft and ran the research lab in Beijing
00:57:06
and he helped create Bing. But then he went with Jason
00:57:08
Kilar and he Co founded Hulu. He was the head of technology
00:57:11
and product for Hulu. So we both spent time in the
00:57:14
streaming wars and then was at Kleiner Perkins and Meta and
00:57:18
other places before the start up.
00:57:19
So Eric shared a passion with our team that what he called
00:57:23
Yahoo's toys would be an amazing set of data to work with to
00:57:28
create a product like this. And so we went to buy his
00:57:32
company last summer and we put him in charge of consumer
00:57:34
search. And so early August last year we
00:57:37
started building. And again, if we go back to why
00:57:41
we did it, we have the third most search volume.
00:57:44
It's small compared to Google, but still a very big business
00:57:47
for us. We had to have the best AI
00:57:50
search results or answers that we could possibly have.
00:57:54
If we didn't we would lose users cuz search is evolving and
00:57:57
changing and if it made sense we would have outsourced it but it
00:58:01
actually made the most sense to to build it.
00:58:03
How much of yahoo.com are you funneling people?
00:58:05
Like are you? Nothing so far.
00:58:07
Why not like? Because we are in beta and we're
00:58:12
a, you know. Cost.
00:58:13
Does it cost more money to distribute?
00:58:15
We're a little more than a month and 1/2 in and yeah, there is a
00:58:18
cost. There's a there's a cost both to
00:58:20
grounding a Bing and we work with a very lightweight model
00:58:23
the anthropic makes called Haiku.
00:58:26
So it's super fast and it, it essentially processes our data.
00:58:30
It's we're not, you know, using, you know, their, their answers.
00:58:33
It's our, it's, it's ours running through their
00:58:36
lightweight model that comes back to us to render in our kind
00:58:40
of, you know, we love our UI. We, we think it's really rich UI
00:58:43
that has some, some personality and some great answers.
00:58:47
So all that together does cost some money.
00:58:49
That cost is coming down over time.
00:58:51
And and then again, over time, I think search advertising will
00:58:56
come back into this category and be able to, you know, make this
00:58:59
really profitable. So that's, but in this early
00:59:02
stage, we are trying to hone it. We actually have a Slack channel
00:59:06
called improve Scout and, and hundreds of Yahoo employees are
00:59:10
in there all the time, like feeding back on what's
00:59:13
happening. We're getting user data in from
00:59:15
people coming through it. I mean, it's in the millions of
00:59:17
people who are, you know, that have been through this and are
00:59:20
using it. But, you know, at the right
00:59:24
time, we'll come out of out of beta and want to unleash it
00:59:27
across all 700 million users. But we just, we're not in a huge
00:59:31
hurry to do that yet. Well, I'm cheering for the
00:59:33
vision of getting traffic back out to the publishers.
00:59:36
Yeah, man, great to hear it. Thank you so much for coming on
00:59:38
the Newcomer podcast. Thanks, Eric.
00:59:40
Great, thanks for sticking around to the end.
00:59:43
Please. If you've made it this far,
00:59:45
you've got to like comment, subscribe, excited to grow the
00:59:48
channel. And of course, you can find our
00:59:50
writing and reporting on sarps and venture capital at
00:59:53
newcomer.co. We also host events.
00:59:57
You can check out what we're doing at Newcomer dot events,
00:59:59
share their podcasts, help us get distribution, support the
01:00:03
channel, comment support, like comment, subscribe.
01:00:06
You know, you know the deal. Thanks.
01:00:08
Thanks for being on the journey with us.
01:00:09
All right, see you next week.