Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - aaron

Pages: [1] 2 3 ... 5
1
Water Cooler / Re: I kinda miss those 5000 search engines.
« on: May 23, 2018, 11:34:51 AM »
I kinda miss submitting sites to 1,500 to 5000 "search engines".  My email inbox used to be jammed with spam asking me to submit my sites or buy some software that would submit for me.  It has been a long time since the last one of those emails.
Cryptocoin emails have sort of taken over that roll.

I also miss directory optimization: where we would craft a title and description for Yahoo, dmoz and Looksmart that contained the keywords we needed without sounding or looking commercial or spammy and hope it got past the editors.
Being a Zeal directory editor (which fed into LookSmart, which fed into MSN Search) was fantastic for instant editing your rankings in MSN Search back in the day.

The whole biz model of web directories sort of died off. It's hard to compete with Google on relevancy unless you are very tightly focused & it is hard to generate as much revenue per visitor as Google does in the key commercial areas. The general perception of the broader industry is links from general web directories don't have value unless they are for local search & have NAP data. In addition most directory sites not only can't draw repeat visitors, but they typically can't even get the category pages well indexed in Google. On the off chance they do get deeply indexed they'll likely get hit by the Panda algorithm. Even DMOZ cratered.

I don't think they are actively accepting submissions yet, but Curlie is allegedly to become a rebirth of DMOZ. One thing I think is a mistake is they put AdSense ads on their help forum before opening the new directory up to submissions. And since their site mostly caters to webmasters they probably don't make much from the display ads on Resource Zone.

But I would kind of like to at least have the option to submit my URL's to a few (maybe 5, please) major search engines besides Google.  That would be rip snorting good fun.
I wonder if a new search engine would be able to scale up to create a competitive product with a strong algorithm & a fairly comprehensive web index without getting blocked by firewalls for excessive crawling.

When overseas many websites won't load due to firewall restrictions unless one uses a VPN or some such.

I also think modern search engines use a lot of the end user engagement data for ranking refinement. Competing on the usage data front would be quite hard. Google owns well over 90% of mobile search via owning Android & paying Apple to be the default search engine. And on desktop devices Chrome has massive marketshare, and I believe Google pays Opera, Mozilla Firefox & Vivaldi for default placement. Bing is aggressively pushed by Microsoft in Skype & Edge, but both Yahoo! and Bing are so ad dense they've seen sharp drops in organic search clicks.

Google's dominance is so strong I saw a Microsoft bundleware option where they were trying to install Chromium optimized for Bing browser. The crazy thing there is they are only one Chromium "security" update away from that being reset to promoting Google.

Most web browsers are also moving away from supporting toolbars.

Oh and I miss Searchhippo.  hhh 

I actually used the HipRank API in one of my directories.  Sites were listed  in the categories by HipRank order.  Well use of the API was free.  It was my little joke on the Google Directory, but nobody understood it (plus it wasn't very good) so I eventually ditched it.  :D
Gigablast is still around & they have an API http://gigablast.com/api.html

If Marissa Meyer had been able to get Yahoo out of their contract with Bing and start their own spidering engine again I had hoped they would revive the AltaVista name.
She did get the agreement amended to where they are able to show ads from Bing, Google or Yahoo!. Yahoo only has to use Bing for (I think it was something like) 51% of desktop searches & gets paid the vast majority of the revenue from the ads (something like 93% without having to incur any of the costs with running a search engine).

But they saw from Gemini data they weren't really gaining ground creating a third parallel network against the first which was already so dominant.

It is also hard to argue that organics would be a differentiating factor for them given how heavy the ad load is on their search results page. Sometimes they have graphical mini banners for other commercial search queries on noncommercial search results. Even when you search for a fairly explicit search topic where they should be able to do good from one of their vertical search services they punt. A few days ago I went to their homepage to search for "Dollar General stock" and there were 4 or 5 ads above the organics, then a local search result panel with map, then the mini insert for their stock price data. I searched again just now and there was only 1 ad above the organics, but sometimes they really weigh the page down with ads.

RKG mentioned in their most recent quarterly report that Yahoo & Bing had their fastest ad click growth in a long time, but that came at the cost of seeing sharp declines in organic clicks.
Quote from: RKG report
While Google continues to dominate paid search traffic with 90% click share, Bing Ads and Yahoo Gemini spend have actually increased faster Y/Y than Google in each of the last two quarters.

The growth of paid clicks on Bing and Yahoo appears to have hurt organic search traffic driven by the two engines, as both posted their largest year-over-year (Y/Y) organic visit declines in the history of this report.

While organic search visits driven by Google were roughly flat Y/Y in Q1 2018, Yahoo organic search visits fell 36% and Bing visits fell 20%. Yahoo and Bingís declines were the largest observed in the history of this report and occurred at the same time that click growth for Bing and Yahoo paid search ads is running near multi-year highs.

Infoseek got borged by the Mouse, but the name has been out of circulation so long only us geezers remember it so you could reuse it and all the kids would think it is new and shiny.

Excite and Lycos are still trying to be portals.  To bad somebody does not try something new in search with them.
If you enter a search into the Lycos search box it will say "There were no results for your search query." Excite is owned by an IAC subsidiary.

I like the new design of HotBot.

Someone tried reviving the Mamma.com domain, but it got hit with an algorithmic penalty as everything on their site seems to rank page 4 or worse in Google.

Many long existing businesses like Cafepress & Overstock.com have been nailed by Google ranking shifts over the past year or two.

Reviving a dead brand might be even harder in the current algorithm given that many active businesses are still getting nailed without going through a shut down period.

CircuitCity.com was relaunched out of bankruptcy once already, but the company which bought the domain more recently now has it redirecting to circuitcitycorporation.com.

Excite, Altavista, LookSmart, Infoseek, Lycos, WebCrawler

Shockingly, half of those or more are still out there and returning search results.
Most search engines beyond Google & Bing are to a large degree retreads of one of those two (at least outside of a few markets like China, Russia & South Korea).

And then some of them are heavily wrapped in ads for one another as the traffic is arbitraged back and forth. IAC still does quite a bit of arb & they're still using domains they've "retired" from an SEO perspective to drive SEM campaigns.

DuckDuckGo has a clean interface.
StartPage.com pitches being privacy focused.
Ecosia plants trees with some portion of your search ad clicks (feeling green, I might just go look for some auto insurance quotes and NYC attorneys real quick :)).

2
Web Development / Re: Wikipedia Link Previews
« on: May 23, 2018, 11:10:03 AM »
So then 72% of traffic comes from outside the site - search or referral traffic. That's higher than I would have guessed given how often I go down the rabbit hole
A lot of people just want a quick answer. Of course Google is trying to scrape the value of those sorts of visitors away via the knowledge graph & featured snippets, but that site ranks for so many keywords.

And for many people going back to Google & doing another search again is easier than using the internal navigation of most websites, particularly for information-dense pages with dozens to hundreds of links on them.

As mainstream media sites erect paywalls, for informational search queries people will have even greater incentive to start at Google rather than follow internal links, to both save their allotted pageviews & to be able to click on another listing if they go beyond the monthly pageviews for that publisher.

3
Marketing / Re: Adobe to buy Magento Commerce for $1.68 billion
« on: May 23, 2018, 08:09:29 AM »
Ben Thompson's Stratechery mentioned in a recently daily update he thought Adobe's acquisition of Omniture long ago was a mistake (due to Google & Facebook owning the online ad market) and they've had to sort of paper over it by continuously buying more & more of what eventually amounted to an entire ad tech stack. Owning the underlying OS for the merchant is sort of the logical end game from there.

With GDPR active in a couple days & Amazon aggressively entering the ad network game to display ads across the web, this is perhaps one of the better ways to try to stay competitive in light of an increasing share of ecommerce-related ads being driven by ad retargeting. For broad-based ad networks which are not brand-driven, owning the cookie is going to be vital to staying competitive with the duopoly & the growth of Amazon. Hosting the content gives one a leg up on the attribution front when multiple ad networks want to claim they drove a particular conversion.

It might be sunk cost fallacy, but not trying to keep that piece of the business competitive could harm the narrative behind the stock. And loss of narrative control tied to questioning management investments is probably not so good when a stock's P/E is close to 60 and the market cap is almost $120 billion.

4
Water Cooler / Re: Remember when times were just a bit SLOWER ???
« on: May 21, 2018, 03:41:30 PM »
Amazing how fast Twitch has grown. Nearly 1/4th the size of YouTube in spite of catering to a single category & not having the decade plus of searchable archives algorithmically boosted by Google ranking integrations, not having the deals with record labels, etc.

Google was pouring 10s of millions into a YouTube creator roll up network 6 years ago
https://techcrunch.com/2012/05/21/machinima-youtube-35m/
They really screwed up letting Amazon buy Twitch so cheap.

5
The research paper does not discuss the ethical issues.

None applicable, due to their new golden rule: "One says a recent justification for moving ahead amounted to: If we donít do this, a less-scrupulous rival will."
https://www.bloomberg.com/news/articles/2018-05-14/inside-google-a-debate-rages-should-it-sell-artificial-intelligence-to-the-military

6
http://www.spyonweb.com/ is a free one specifically for tying together sites under a common AdSense or Google Analytics account, though there are other more advanced tools like https://nerdydata.com/ & https://publicwww.com/

7
Water Cooler / Re: Don't get distracted
« on: May 16, 2018, 12:38:28 AM »
In the context of this thread, the following - from our future AI overlords at Google - is easily one of the most heartwarming lines I have read in years.
https://www.bloomberg.com/news/articles/2018-05-14/inside-google-a-debate-rages-should-it-sell-artificial-intelligence-to-the-military
Quote
One says a recent justification for moving ahead amounted to: If we donít do this, a less-scrupulous rival will.
So the only way that logic stops working is to become the singular least-scrupulous entity in the world.

Talk about justifying a race to the bottom!

8
Water Cooler / Re: YOUTUBE music thread
« on: May 15, 2018, 01:39:17 PM »
https://www.youtube.com/watch?v=GCa_TKn9ghI

R‹F‹S Del Sol - Tonight

9
Monetization / Re: Amazon halts Google shopping ads
« on: May 12, 2018, 07:50:45 AM »
About a year back RKG noted Amazon was more competitive on mobile than desktop for PLA distribution ad.
https://www.merkleinc.com/blog/amazon-focusing-mobile-google-product-listing-ads
So many consumers already have Amazon.com accounts coupled with stored payment information and 1-click ordering meant Amazon gets a far higher conversion rate on mobile than most advertisers do.

When the Walmart & Google relationship sours Google will likely have to buy Target to try to stay competitive with Amazon.com.

10
The trend matters as much as the rate because ultimately the trend points to where rates will be unless there is some counteracting force. The budget cuts eventually end up causing crime epidemics. As criminals get away with criminal activity it emboldens more criminal activity.
https://www.london.gov.uk/press-releases/mayoral/budget-confirms-110m-for-met-police
Quote
In recent years, the Met Police have had to find roughly £600m of savings and the Mayor has found a further £150million of savings since he took office. This has led to the loss of a third of police staff posts, which are down from 14,330 to 9,985, as well as two-thirds of police community support officer (PSCO) posts, which are down from 4,607 to 1,591. In addition, there are now 114 fewer police station front counters and 120 fewer police buildings.

https://www.thesun.co.uk/news/politics/4969204/theresa-may-slashes-413million-from-police-budget-after-vowing-to-protect-cops/
Quote
THERESA May has secretly slashed £413million from the police budget - despite vowing to protect cops, it's been revealed.

The shocking figures were released on the eve of the Chancellor Philip Hammond's Budget speech, which he'll deliver at the House of Commons today

https://www.theguardian.com/uk-news/2018/mar/20/theresa-may-accused-of-misleading-public-over-extra-police-funding
Quote
Theresa May has been officially rebuked for misleading MPs and the public over false claims that the government is providing an extra £450m in funding to local police forces in 2018/19.

The chair of the UK Statistics Authority, Sir David Norgrove, ruled on Tuesday that the claim made by May repeatedly at prime minister’s questions last month “could have led the public to conclude incorrectly” that the government was providing an extra £450m for police spending over the next financial year.

11
Map understates aggregate crimes as well. Pretty sure my car was parked in that one part of the map that wasn't red when someone did a smash & grab to steal my friend's luggage at around noon. Lots of such crimes go unreported because people know nothing will come of them. I think my friend who's luggage was stolen did take the time to report the crime, but he also had to buy a bunch of new clothes for the rest of his ruined business trip after his luggage was stolen & I had to buy a new window.

12
Let's talk about the cost of paying a living wage to reporters, support staff, secretaries, janitors, security guards, building rental, travel costs, etcetera...

The only people who start at a position of higher margins are parasites like Huffington Post who hire millenials to paraphrase the work of legitimate news organizations.
Crazy thing is I don't think HuffPo was particularly profitable. They were sold on the narrative of "growth" but then that growth sort of went away in part due to them shutting down their contributor network (perhaps in part for some combination of improved reach score on social channels, editorial overview cost savings, advertiser safety differentiation to lift CPMs above other race to the bottom channels).
http://adage.com/article/media/huffington-post-broke-146-million-revenue/299293/
Quote
So if The Huffington Post -- which is 10-years-old, hauls in more than 200 million unique visitors a month and cranks out roughly 1,200 posts daily on the backs of reportedly poorly paid or unpaid writers -- can't turn a profit on $146 million in revenue, then how are the other, venture-capital fueled sites with smaller audiences and fewer relationships with advertisers supposed to achieve profitability?
https://www.nytimes.com/2015/07/05/magazine/arianna-huffingtons-improbable-insatiable-content-machine.html
Quote
The Huffington Post broke even last year on $146 million in revenue, according to someone familiar with the siteís finances ó is considered a temporary problem that will eventually be fixed by the sheer size of the readership.
https://www.vanityfair.com/news/2016/09/why-arianna-huffington-left-the-huffington-post
Quote
in a few short years, the site was experiencing a Digital Age version of a midlife crisis. It was reaching 26 million unique visitors per month, an astonishing number, but in the Internet business, sites either grow or shrink. And to grow, the Huffington Post needed more money.
...
the very year Huffington cut the deal with Armstrong, 2011, turned out to be the publicationís only substantially profitable year.


Ultimately so long as the market has attention dominated by a few platform monopolies which keep growing more dominant by the day, copy external innovations into their core products (Snap as Facebook Labs), buy services to track growth of competitors (like Facebook buying a VPN), etc. ... there will always be a few interns willing to eat on credit from their parents & illustrious opportunity points. And there'll be a few VC-funded media innovation plays that sell "a new way" & growth story to old school outfits.

But for those who long for something more than that, I think a shift to subscriptions will be needed for publications which aim to be much larger than an individual or small team who covers a specific niche or vertical.

And I think reformatting across proprietary formats (Instant Articles, AMP, having Medium as the CMS) is going to be a mistake for most publishers who want to sell subscription access.
http://www.niemanlab.org/2018/05/medium-abruptly-cancels-the-membership-programs-of-its-21-remaining-publisher-partners/
Quote
After each strategy shift at Medium, partner publishers have said they felt the rug being pulled out from under their feet, with no notice. In the spring of 2016, Medium rolled out a suite of tools for publishers and attracted big names like Bill Simmonsí The Ringer and California magazine Pacific Standard, along with beloved indie outlets like The Awl Network, Film School Rejects, and Femsplain. All of these publishers have migrated off Medium (most depressingly, The Awl and Hairpin arenít around anymore at all).

That's a good observation. Yahoo News does things similar to that.

The problem with that approach, is that you run the risk of spiraling downwards to the lowest common denominator, like Google. Google aims to please so they show the most popular content. Which is why there is so much Kardashian and Taylor Swift bullshit in Google News than you might find on the NYTimes.
It is quite easy to compare the click score for the individual vs the click score for the general population to prevent the race-to-the-bottom rewrites of the latest Kardashian news. Stories could also be topically scored not only by section but through tagging, the section they are in & through semantic analysis.

And if one is selling a subscription there is greater alignment with not going through a race to the bottom, so maybe those articles are not only not promoted algorithmically but perhaps they are not published in the first place.

I think the harder problem (particularly with open networks) is not over-personalizing the way YouTube does, where one starts with a video a little bit out there & a few autoplays later is deep in crazy conspiracy land. But, again, that problem goes away with a paid subscription business model driving access to editorially approved content.

Of course there are different points of leverage for aggregators vs individual publishers. Wherever there are ad-funded publishers in a competitive channel there will be incentive to game the system. And even subscription services have an incentive to game the system if things are split between a central network and individual authors. There were stories about spam auto-gen Kindle publishers that would have people open page 2,792 or such on the book to get a large share of reader subscription revenues.

13
Not sure about the race to the top concept as they've also been known as the city with half a police force & have published lists of crimes they wouldn't respond to.
https://www.nbcbayarea.com/news/local/Suffer-These-Crimes-in-Oakland-Dont-Call-the-Cops-98266509.html
https://www.sfgate.com/crime/article/Oakland-cops-not-coming-unless-crime-is-violent-3181544.php

Lots of random junk stops, but not so much actual policing
https://www.eastbayexpress.com/oakland/why-oakland-police-cant-solve-crime/Content?oid=4288948
Quote
In Oakland, investigating and solving crime hasn't been a priority for decades ó and perhaps never was. In a memo last week to the Oakland City Council, Police Chief Sean Whent reported once again that OPD is failing to solve the vast majority of crimes in the city. But it's worse than that. Oakland police detectives, When revealed, don't even attempt to investigate most crimes ó other than homicides and sexual assaults.
Which is part of the reason you can park in a nice area of town for lunch with a friend who came to town to find that around high noon your window was smashed so someone could steal his luggage. Always use the trunk!

14
Worth noting they announced it is something to come rather than something which was immediately implemented.

Some of the YouTube vids for that stuff are extremely cringeworthy. A talking headshot of an MLM-styled guy reading a scrolling salesletter for some coin he gets a commission on for getting you to sign up for. Ads to rent out your spare computing power where you are almost certainly paid less than the amount of electricity they use as they run your computer hot cryptomining, etc.

Also in some other banned categories they have later re-activated them.

Bails bonds recently nuked, just like payday loans were a while back
https://adwords.googleblog.com/2018/05/google-bans-ads-for-bail-bonds-services.html
(though payday ads only nuked in certain locations, showing it is a political choice spun as a moral choice, especially when combined with Google's LendUp investment)

recovery ads re-enabled 3 months after they were nuked
https://www.theverge.com/2018/1/13/16887898/google-adwords-pulled-rehab-addition-center-global
https://www.theverge.com/2018/4/16/17244192/google-advertising-drug-addiction-treatment-centers-abuse

The binary approach they take in many categories has to be tough for businesses in those categories.

The big issue with banning the ads is they get no revenues while diverting the traffic to organic listings (subsidizing rather than defunding the competing channel). You'd think in categories they ban ads on they'd at some point put some large & ugly graphical sort of PSA at the top of the page to push the results below the fold & to guide users to parallel search queries where the ads are not yet banned.

15
If publishers have a direct subscription relationship with the reader they get less out of print distribution than they do online.

The marginal cost of distribution online is effectively zero, so they start from a position of higher margins.

In addition, online they can track what people read & are interested in, highlighting more related articles. They can also see when a person's usage has dropped and work on re-engagement campaigns where they promote articles they think that person would be interested in across the web, perhaps personalized to an audience of 1.

If they send out a print issue there's no tracking of user engagement & it is harder to cross sell events or sister publications & such in breaking news announcements (in addition to the paper being slower to publish new news).

Pages: [1] 2 3 ... 5