Tumblr’s stumbles under Yahoo
mashable.com
mashable.com
I contacted their ad team and was told the minimum ad buy was $25k. I explained the situation and said that given the popularity of our product among the Tumblr community, I had no doubt we could find a revenue-positive way to advertise on Tumblr, but that we couldn't commit to $25k sight-unseen. I never heard back from them.
Considering how easy it is to advertise on Google or FB with just a couple bucks, I was shocked that Tumblr is so inflexible. I guess it's not surprising they couldn't get to $100M in ad revenue, since they'll only talk to companies that can afford $25k experiments.
With big media properties that are being cleaned up, a lot of that early focus is typically on big brands who are doing large branding efforts that have cool and sexy creative and are super customized to the platform. They want to reach lots of people, and the property in turn wants to leverage it as a large case study to attract other brand advertisers.
What you were looking for would not help meet their objectives because your budget was likely too small, and because they are a new display-driven property, they don't want advertisers focused on performance. Why? Well, most performance advertisers are not super savvy with how they look at display performance, and would be looking at it from a last click measurement standpoint where it would almost certainly bomb (like most display, social, etc.). Big brands doing branding plays are looking for reach and engagement (or post-impression performance metrics in some cases). This is better for Tumblr for purposes of building case studies and selling in to other big brands, because big brands tend to not care as much about overpriced CPMs, which means more margin for Tumblr.
Again, down the line it might have made sense that you could make a lot of money for your brand and might have been a good fit, but realistically, you just weren't their target. Same reason FB weeded out all the crappy affiliate ads to the best of their ability and started pushing big brands for engagement-based stuff.
This sounds like an effective way for any company dependent on sales to fail. Especially online advertising sales where your incremental cost is approximately zero.
There should be a term for management that makes this mistake.
With small budget direct response advertisers, all too often you get people who promise they will spend more if it is profitable, but because they don't know how to properly value display (see my last click comment) they won't see last click success, and will thus not be a repeat advertiser. Beyond that, they will also contribute to spreading the word that Tumblr is not a quality inventory source, which could kill them before they had a chance to really shine.
If I'm trying to grow ad sales on a property as large as Tumblr with the reputation it had, the first thing I'm going to do is make it clear that the focus is first and foremost on big brands and not the little guy. That's what brings in the dollars, not the latter. The little guys come in once you are already at scale with proven success for big brands.
Once that is successful with your limited sales resources, then and only then do you create a self-serve platform for small direct response advertisers with minimal 1:1 contact (ala AdWords, Bing Ads, etc.).
You need to think of the long game and think of where you would prioritize your resources. Would you rather get a small inconsistent cashflow of people who are likely to talk crap about your inventory because they don't know how to properly measure it if they don't see a positive ROI after spending a couple grand? Or would you rather focus all your efforts on big brands who know what they are doing, have the budget to do highly visible tests, and aren't afraid to "waste" 10's of thousands of dollars testing something new that might not work out?
Lastly, I should have said "profitably", not "revenue positive". Thanks for not dinging my accidental misuse.
It isn't primarily about the time cost for the sales team. It is about what sort of image they want to portray for "what companies advertise on Tumblr." If you see a bunch of acai and weightloss ads, and you're a big brand, you're not going to advertise there. That's a problem, because odds are you have a lot more money to spend. So in order to attract the right advertisers, Tumblr has to portray an image of success that aligns with those targets.
So your statement about small and midsize startups doesn't make sense in that context. They don't want small and midsize startups. Again, they want big brands. And you approach the business strategy for that very differently.
Also, "native" ads aren't really that different than display. I'd personally consider it a subset of display (like video). You generally are looking at post-impression (vs. click) performance, reach, engagement, etc. It is VERY hard to measure well. So small and midsize startups are again not the right fit if they care about measurable ROI.
Tumblr doesn't want "some of these companies" that might grow into bigger companies. They want the big companies that have the big dollars and will pay high-margin CPMs TODAY.
Once you have that business funnel doing well, THEN maybe you consider expanding out to a self-serve platform with lots of small guys. Because at that point large advertisers will know you are a solid inventory source and not be swayed when a tiny little startup who doesn't even know how to set a conversion tag claims that the clicks they are tracking aren't driving them lots of profit (see pretty much every post griping about FB ads on HN).
Also, setting up a self-serve system takes a LOT of work and resources compared to tossing together some janky (but functional) tools that your internal ad ops team can use to traffic things. Which again makes it low on the list.
And "profitably" vs. "revenue positive" doesn't matter--if you care about direct response performance, they don't want you as an advertiser at that stage.
Again, not saying you or other smaller companies couldn't find success on Tumblr. But just because you want them does not mean they want you (at least at the stage they are at). That $25k min. is to weed out smaller budgets and direct response advertisers who need to test small and see profit before continuing to spend more.
For those out of the loop: http://imgur.com/gallery/jKs0n
Although it wasn't $25K, it was closer to $18K. After grilling the person over email over basic information, they couldn't give me a clear strategy on how my money would be used, the ads constructed, or where the ads would be placed.
It just felt like they were extorted money from companies claiming we needed them not the other way around. It was eye opening they weren't interested in gaining some recurring revenue as opposed to try and get the most money possible in one fell swoop. It was really perplexing to me.
Of course the strategy flopped, but that's another story.
Another factor is the failure of Tumblr's staff to appreciate the diversity of ways Tumblr can be used: as a picture blog, as a server status log, as a way of communicating person to person, as a way of communicating person to corporation, as a way of creating funny writing projects or online roleplaying, or just as a content aggregator. All of these are valuable uses of the platform. Not all of them are ever accepted or catered to, in fact, most of them are ignored.
I'd like to see more on why Tumblr's dev team does what it does than on why its ads team sucks.
- search is actually usable now, was completely useless before acquisition
- new messaging features are better than stuff they had before
- rich text editor better than the one they had before (except if you want to edit in HTML or Markdown in which case it completely sucks now, but very few people probably post in HTML or Markdown)
- replies, I have hardly ever seen them used before they were removed. in any case, the new implementation is much better
- should I even mention the freaking video player?
Of course the users will complain, but I mean this is Tumblr, the userbase will complain about everything.
Replies were HEAVILY used by my part of the Tumblr community, and the changes to them really screwed things up for us. Like I said, lots of use cases for the site.
I certainly don't feel on the whole that tumblr was better when I stopped using it than it was when I started. Even if that is true now, it wasn't true for a long time, so my argument remains the same: its decline is due to inadequate tech work for a long period of time.
How are so you sure that a feature you never used is better than it was before it changed? Even if I'd never used tumblr, I would find this response to be incredible, in the original sense of the word.
Replies are so bad that I don't use tumblr anymore. They used to just load in a normal HTML page, but now you get a page with lots of suggested posts, and some sort of pop-over with replies that's too slow to really be usable even on my brand new iPhone. It's usable on desktop if you want to wait a few seconds for enough CSS and javascript to load that the pop-over shows up, and then scroll through and read a long list of replies.
These are not inherently evil people, but they were allowed to impose terrible, deep mistakes on the staff and users for a very long time. No one was in a position to stop the unforced errors, not for years. Which is why Yahoo! was a fitting acquirer. (The sale to Yahoo! probably prevented the platform from foundering)
The core product is still extremely useful to the web (especially because it's free). The demand for such services, and a capital market that rewards any kind of operation as long as it's acquiring ~u s e r s i n t h e d e m o~ - it gives a very long runway to digital publishers like these.
But 9-10 years is a long time to be around without a stable business model. And, as a very worrisome problem, you would not be able to persist your user data or your established URLs if the platform foundered. A shutdown of Tumblr would be a terrible single-day source of Internet linkrot.
But, same for Flickr, which is a much more admirable (and longer-lasting) product run by capable people but faces the same existential threat. Sad, because Flickr should have been Facebook (eventually).
I don't know. Flickr is optimized around sharing photos (at full resolution if desired) to a broad audience. There is a largely optional community angle but a lot of us barely participate in that.
I have a short list of things that I wish flickr did better and there are doubtless things it could have evolved into without harming the store/display-photos use case. But, although it's been pretty stagnant I'm honestly not sure what more I really want it to do.
I agree it would be very unfortunate if flickr were to either shutdown or do something like turn off non-paid accounts. I find it a useful source of CC images and I am contacted from time to time by people wanting to use photos of mine.
Just remember where it was when Yahoo acquired it.. $13m in revenue and $25m in expenses for 2012 (net loss of $12m for the year). With $125m raised, and $114m of debt on the books.
Yahoo acquired them just a few months after they announced these results for 2012.
It's easy to blame Yahoo for Tumblr's problems.. but Tumblr wasn't Twitter or Instagram. It was a tiny business (in comparison).. that without Yahoo very well could have been out of business by now.
http://www.businessinsider.com/heres-tumblrs-total-revenue-f...
Yahoo knew this, and wanted this demographic for their ad ecosystem. They often thought of themselves as a 'media company', and they liked the creative aspects of the community. They also, wisely, knew that Tumblr users would never accept being forced into using a Yahoo account, so they had to keep the userbases separate.
But this meant they couldn't use Tumblr to build Yahoo's social presence and its brand. They'd harvest some analytics, display some very loosely targeted ads to people highly averse to ads, but that mercantilist relationship was all was to be. In that light, they clearly overspent.
Stating something as fact (oh, a source disputed this), artful use of connotation ("a yahoo source played it off..."), and lastly (where I stopped reading), calling the founder of Tumblr s "20-something". You don't get to call a 29 year old a "20-something".
Feels overly sensationalized, even if yahoo is where start ups go to die.
I don't think there's any need to add an adjective to it unless the intention is to convey ~"too young; naive; inexperienced" as seems the case here. They could simply say e.g. "John Doe, 29, [...]".
Y'all should be cognizant that there's a human being with their own priorities writing articles liked this.
My point in showing how biased this article is so you can remember to take it with a huge grain of sale. Probably less story here than it makes it out to be:
Tumblr was failing to execute on the revenue side of their business, both before their acquisition, and then after.
At some point, Tumblr had to try and become a business and not just a website.
This article points blame at Marissa, but shouldn't it really be pointing at David, or a team of leaders?
If you hire a roofing company to fix your roof, and one of the workers screws up, it's still the responsibility of the owner who hired them.
This line stuck out to me as very true reflection of what the Valley sometimes looks like:
> Tumblr’s stumbles under Yahoo may go down as a cautionary tale, both for the perils of a large corporation buying a hot startup and for Silicon Valley’s belief that any social network reaching hundreds of millions of people will inevitably generate boatloads of cash one day.
There was talk of "We'll make it an advertising platform" but did anyone really ask the Tumblr team "How" ? (I'd love to hear info from any insiders on HN)
Makes me wonder how many times a day the execs at Facebook wonder how they can ever monetize expensive companies like Whatsapp.
Not everybody uses their Facebook account but they may be more prolific on Instagram. They might not like being constantly monitored by Facebook Messenger but even with WhatsApp's end-to-end encryption, Facebook is still provided with valuable metadata, and the more diverse data sources, the more metadata resembles regular, easily interpretable data.
It's like Microsoft buying LinkedIn: the market was clearly bearish on the company, but a tech Titan like Microsoft has a lot to gain by keeping its finger on the pulse of potential hires around the world - with a crucial head-start before headhunters can find star performers.
Some historians of the automobile industry (Paul Ingrassia) suggest that Oldsmobile would have gone much further if it had remained an independent company. Wikipedia points out it was the #1 car company, shortly before it was bought:
"In 1901, the company produced 425 cars, making it the first high-volume gasoline-powered automobile manufacturer. (Electric car manufacturers such as Columbia Electric and steam powered car manufacturers such as Locomobile had higher volumes a few years earlier). Oldsmobile became the top selling car company in the United States for a few years around 1903-4... General Motors purchased the company in 1908."
https://en.wikipedia.org/wiki/Oldsmobile
John F. Love points out that in mid 1960s Burger King was building twice as many restaurants as McDonalds, and was on track to surpass McDonalds by the end of the 60s. At that time, Burger King had very smart management. However, Burger King was then bought by Hormel, who regarded it as simply a place to dump surplus beef that couldn't be sold through other channels. And that is why McDonalds is the #1 burger company.
The software industry is full of such stories. Roomkey.com (it's pre-purchase name was Hotelicopter) might have transformed the hotel industry, but it was bought by a joint venture of Choice Hotels, Hilton Worldwide, Hyatt Hotels Corporation, InterContinental Hotels Group, Marriott International, and Wyndham Worldwide. This joint venture regards Roomkey merely as a bargaining chip to get better rates from Expedia ("Give us better rates or we will invest more in Roomkey".)
I could try to list all the innovative startups that Cisco has bought, cannibalized, and then discarded, but it would take me all day.
Mergers are the main way that innovation is destroyed in our society. Over and over again small startups are bought, and then killed by the different needs of the new owners.
Edit: Android might be an example acquisition success story.
[1] http://www.businessinsider.com/2008/7/twitter-buys-summize-f...
Off the top of my head examples:
* Google Maps (Where 2 Technologies)
* Siri (Siri Inc)
* Visio (Visio Inc)
* Firefox (born of post-acquisition AoL money)
It didn't occur to me that millennials could be oblivious to the existence of Yahoo. (Afterall, millions of people still use Yahoo Mail, Flickr, etc.)
> Karp, a coder at heart, is described by colleagues as “openly hostile” to traditional advertising.
I suspect Tumblr users, much like Reddiors, are allergic to advertisements.
> Traditional marketers... were also skittish about some of the raunchier user-generated content on Tumblr.
No shit. In many respects, Tumblr is essentially a bootleg porn site. From Day One, people have cited Tumblr's reliance on "adult content" as a major problem for Yahoo. I thought the would just ban adult content, but then I realized that they can't because they would lose a significant chunk of their userbase.
Same thing with Twitter IMO. Everyone hates promoted tweets, but many love Twitter. In particular, it's great for political dissent and organizing protests. Good luck monetizing that.
In the case of SnapChat, tons of their most active and loyal users are in elementary / middle / high school. I was in a restaurant a few days ago an saw 4 children whose shoulders came just above the table sitting there for an hour snapchatting each other, and suddenly the massive user base of it made sense. Again here, good luck extracting money from children inside an app that they very likely try to hide from their parents.
In the case of FB Messenger / WhatsApp / Telegram / Line / Kik / Viber / Kakao: this stuff is commodity chat, and the value lies in other people in your network using it. If it's found that any one of them are doing something that users don't like (like showing ads or reading your messages or mining who you're talking to), the users will have no problem finding alternatives. It's like trying to monetize Skype, without the value proposition of being able to call land lines. It's like saying that AOL Instant Messenger was worth $1bn. The value is in the utility of the thing, which crucially, is trivially replaced by a competitor. At least Facebook has the benefit of not requiring returns on messaging and instead pitching it as an integration with their core platform, which is great at getting ads in front of eyeballs. Again here, users love these utilities, but I don't ultimately think that they're economically viable businesses.
Also, not sure what your point is with regards to my comment. Maybe my point wasn't clear. My point was just because something doesn't directly generate profit doesn't mean it's useless, since it can act as a critical strategic asset to another company. Youtube was an excellent acquisition for Google despite the fact that it's not so good in terms of profit. Tumblr could have been something like that for Yahoo, except that there were many things wrong with Yahoo itself and the vision couldn't be realized.