| This 54 message thread spans 2 pages: < < 54 ( 1  ) ||posting off |
|Why no PFI at Google|
what is going on in Google's mind!
| 10:36 am on Feb 2, 2003 (gmt 0)|
Ok timeout here for a minute,
Can someone here (hopefully GoogleGuy) can give me a "GOOD LOGICAL EXPLANATION" on why Google doesn't have a PFI program! because I...just...don't...get it!
Sure you might say that Google is not public like other SEs and doesn't have to satisfy their money hungry shareholders and sure you might say that they have a good free Add URL system that can take your site in their db pretty fast but my answer to this is ....<snip> :-P
Public or not who would be stupid enough to put aside thousands of web site owners here willing to pay at least 39$ (or over 50$ in my case) to include their web sites in Google's db in 2-3 days like the others are doing...
Somebody here is sleeping. It's Either me who is way to blind to see the obvious reason why (i bet i'm the only one here who don't get it) or the CFO at Google is sleeping on the job and are letting go millions of $$$$$$$$$$$$$$$$$ (Hellooooo anybony home) each year in PFI revenues.
Does someone here have the patience to explain me what is going on here because i'm clueless on this one.
| 11:30 am on Feb 11, 2003 (gmt 0)|
Interesting thread and im having a stab at being GOOD and LOGICAL
>>But the beauty of PFI is that you can take that page and modify it so that when the PFI db refreshes in 2 days you stand a chance to rank better and not have to wait for another month or 2 to find out that you moved to the...100th spot! But here again my sites would have to follow the relevancy rules to gain rank! <<
That's the problem. Google does NOT want people modifying their pages all the time to achieve some sort of competitive advantage in rankings. They would prefer to have a system where they ALONE decide relevancy.
Google does not want sites written by SEOs to have a better ranking than say a site that is not SEOd.. eg. government sites with objective info, university research papers, charities, and a heap of sites that have excellent info but will never page to SEO or pay for inclusion.
The final answer is that every page myust have an equal chance of being ranked well, and content should be the only criterion. PFI just makes Google's job that much harder.
Let's face it, the only sites who pay for inclusion are those that have the money to pay and most likely have SEO backup. Almost all are sites which have a revenue or profit marking motive. They are not necesarilly the ones that people want to find in their SERPS.
Now Google offers an option for those that want to get exposure fast. Two actually. Adwords and Sponsored/Premium links. People know that the main index is as unaffected as possible by paying for ranking. They use that to find objective and non-commercial info. They go to Adwords to find things to buy, that people have paid to be exposed in.
Its a simple system.
I think its good, not because of some ethical or emotional attachment to non-paid SERPS, but because it just simply provides the best results for information searches, and provides the draw for people to use google - and then see the paid links on the side and other services than can become subscription. that is their business model.
My advice is if you have a commercial site, PAY for inclusion through Adwords or Sponsored links and forget about spending time on optimizing. It works for us, and can be more cost-effective.
Google will not change as you wish as it invalidates their whole revenue model, so stop dreaming.
There are so many existing and potential revenue streams for Google that i agree with others that say PFI income will be a spit in the ocean, especially if it downgrades their killer loss-leader product.
| 11:56 am on Feb 11, 2003 (gmt 0)|
|Can someone here (hopefully GoogleGuy) can give me a "GOOD LOGICAL EXPLANATION" on why Google doesn't have a PFI program! |
Because, logically, they would also need a PFE -- Pay For Exclusion program.
This is a simple set up whereby they sum the monies received for and against a site each month. If the total "for" cash is positive, the site is in; if not, sorry, no listing.
PFE allows sites to outserp (not outrank -- there is a crucial difference) competitors simply by paying enough to have them removed. This saves a lot of time doing SEO and other hard stuff.
Other SEs have a NPNI -- No Pay, No Inclusion -- program to complement their PFI program. But they haven't thought it through all the way. Where's the logic in PFI and NPNI without PFE?
Google runs on NPIA -- No Pay, Included Anyway -- so it logically follows that it doesn't need PFI, PFE and NPNI.
I hope that's clearer now :)
| 11:57 am on Feb 11, 2003 (gmt 0)|
Okay, here goes:
>>You seem to know a lot about the server power it takes to create and run a PFI program like most of the others SEs are doing here
First, why would Google want to create a PFI like most other SEs?
Before we delve into the math behind it, let's take a look at the anecdotal evidence. The current Google dance takes longer than 2-3 days and that's not counting the deep crawl, which you seem to think happens almost overnight.
Assume there are one hundred thousand bots crawling the web carrying back information, each bot is responsible for 30833.25 pages. If they can grab a page per second it will take each bot 8.56 hours to just grab the pages. Now Google needs to process the information, say they can run a page per second including over 100 variables processed for each page. We're at just over 17 hours now assuming each page takes 2 seconds to process and now it needs to be returned to the index given its new ranking and PR and say that takes a second.
Now we're up to 25 hours and Google has already sent out the next round of bots to keep up with the demand for the PFI folks that need to have that site crawled, processed and returned to the index. That's at a page every 3 seconds.
Of course Google doesn't have to spider the entire index everytime a site is added, they could just spider the new site and follow the links, which is where they run into problems with applying Pagerank. Say that brand new PR7 link you just received is from a site that Google now finds only has a PR6 and isn't a PFI site. They can't give your site the PR7 link and drop the PR of the site that isn't in the PFI scheme, in fact, they can't even determine the PR of the site linking to you without either relying on the old data or respidering and recalculating all the sites that link to yours, and of course, all those sites that link to the sites that link to yours and so on.
Of course you would want the full benefit of having Pagerank applied to all your PFI sites, which of course means that Google has to constantly calculate the new PR value of all the sites linking to you. Then of course there's the question of sites that you link to. Do they receive the benefit of your constantly calculated Pagerank even if they aren't in the PFI scheme?
I may be completely off and Googleguy may have something funny to tell everyone at the 'plex. :)
I do work with mathematicians though and they seem to think that relational database calculations involving complex mathematical algorithms chew up system resources at a very rapid rate. They also insist that Google doesn't have "monster computers". I think the quote was, "Linux Boxes, what? No Crays? How many petaflops are they getting on those bittie boxes?"
| 12:30 pm on Feb 11, 2003 (gmt 0)|
Getting into the database takes a couple days now. Who would pay for something that happens for free? A totally separate issue is recaculating the rankings of the entire jillions of pages on the Internet that are in the permanent database. This is something that is done monthly. Thinking this should be done everytime somebody wants to pay a few bucks to get his geocities page some true pagerank is just absurd.
The day the instant a page is put online the entire universe of webpages will have an immediate and permanent recalculation of pagerank is not around the corner.
| 1:10 pm on Feb 11, 2003 (gmt 0)|
i read somewhere that google use cheap (under $1000) machines using a custom linux os in clusters for spidering, using a very stripped spec, all off the shelf parts. not exactly "monster Unix servers"
if thye did a pfi program, they would have 2 databases, one pfi, one free. digitalghost points out some of the problems with intigrating the two
i can think of some commercial problems too, like how much should google charge for pfi? if they charge high, like $100 per URL, they get slammed for pricing the little guys out of the market, and sliming up to big business. if they charge low, like $15 per url, its basically a spammers backdoor, and they get slammed by everyone when the quality of the result sdrops through the floor
if they have a pfi program, they d*mn well better make sure that submitted pges get spidered. that means having a load of tech support people dedicated 24/7 to monitering problems. thats a cost they dont need, and that eats into the revenue. also they need to make sure that webmasters who haev "why isnt my site in yrt" questions get an answer quickly, and how many manhours a month do you think that will take? that costs too
|Unless (and that to me is THE real reason) Google doesn't want us SEOs to optimize our sites so fast... Things that make you go hmmm isn't it hehe ;-) |
hmmm, do you think that maybe googles search techs are smart enough to change the algo a bit every few days too? just enough to play with your attempts to crack the algo without totally trashing the results? its hard to hit a moving target, fella
| 1:45 pm on Feb 11, 2003 (gmt 0)|
First of all i'd like to apologize for being a bit rude at the beginning with you here, it's my way of attracting people to a thread and have hot debates where good info is shared. Remember my nick, you'll see me shake the boat more than once here ;-)
Ok back to the post now.
You know what, i think that chiyo has put the finger on the reason. I'm not 100% sure but if i had to bet on a reason i would bet on this one.
Here's what he said:
>Let's face it, the only sites who pay for inclusion are >those that have the money to pay and most likely have SEO >backup. Almost all are sites which have a revenue or profit >marking motive. They are not necesarilly the ones that >people want to find in their SERPS.
I always thought that there has to be a better reason then server capabilities or other reasons to not have thought of a PFI program that would have made them literally millions per year and many of you said that relevant listings was the reason but not a good explanation was delivered until chiyo said that. Thanks dude, "now, i think i get it" :-D
I always thought that Google didn't like us (SEOs) to be listed that high whether we pay or not and the fact that by paying we could still fit the relevant filter but beat in "speed" the other sites may be the main factor here.
By being able to modify our site 15 times in the first month it makes the scale moves on our side meaning that a good SEO can outrank any old #1 sites ranked by Google alone and even if we respect their algos they might prefer finding "on their own" that #1 site that has the perfect algos.
In other words, they want our site to be either perfect at the beginning and if not then just too bad hehe :-P. You are left to try optimizing it for the next months...good luck hehe.
Although i'm still confused about the price to pay for that choice losing millions of $ along the way because WE (SEOs) are good...very good and any good SEO can hit the Top 10 after 2 or 3 months by either trial and error or reverse engineering.
So here we go again, does it worth to lose millions because they don't want our highly optimized sites to hit the #1 spot too fast EVEN if our #1 site respect Google relevant rules? Well you tell me hehe
Thanks to everyone. If you have a good theory on that please post it here hehe :-D
| 2:04 pm on Feb 11, 2003 (gmt 0)|
I'm not sure but i don't think that tech support or server maintenance is THE reason here but again i can be wrong here.
The way i see it, 20 million $ each year is a loooooot of money here to hire what? 5, 10 tech support guys? and how many super boosted Cray do you need to crawl just a few thousands of PFI sites here and merging 2 SQL database together? certainly not 20 millions a year right? at least i don't think so.
But that being said, you were right about one thing. It may be hard to hit a moving target but THAT is why (i think) they don't go PFI because they are happy with their ONE UPDATE per month and they would hate resetting their algos every 2 days or so just to avoid us cracking their algos.
I would bet my shirt on that. They...just..don't...want...us to crack their algos and be listed #1 in the first 1-2 months.
Now lets say that you spam your site with kws and Google list you #1 on the first month and i have a site similar to yours but with just the right algos to outrank your site and be a more relevant site for their surfers, shouldn't Google help me outrank your site as soon as possible here instead of letting me wait 2 months to finally outrank you?
Anyway, maybe we'll never know hehe.
| 2:23 pm on Feb 11, 2003 (gmt 0)|
do the numbers
5 x techies @ $50k per year (includes "on call" pay) = $250k
12 x customer support spods @ $20k per year = $240k
16 hours of senior management time per month = 192 hours, worth (conservatively) $20k per hour = £3840k per year
= grand total of $4.33 million
round up to $4.5 million to allow for hardware/software/bandwidth
doesn't look like such a good deal any more really
allowing that hey charged the $50 per url that you said you would pay, they need 90000 urls just to break even
anyone know how many urls are in other pfis, ink, fast etc?
ok, probably they could get 90k urls easy, but i really dont think that theres that much money in it (for google anyway. many would refuse to pay having had bad experience elsewhere, or because they can get in for free after a max of 4 weeks anyway), and it would be a major headache in the first few weeks, and they'd have to do it FOR EVER. i dont think they want to tie themselves down that way
| 2:29 pm on Feb 11, 2003 (gmt 0)|
>>You know what, i think that chiyo has put the finger on the reason. I'm not 100% sure but if i had to bet on a reason i would bet on this one
Your line of reasoning is quite common. :) Basically, you will disagree with any theory except the ones you have already chosen to embrace. The fact that you overlook your own argument countering Chiyo's is humorous though.
We were speaking of PFI, not pay for placement. Google would accept your money, spider the site, determine the ranking and place it accordingly. The 2-3 day time frame versus a 2 month time frame for acceptance is irrelevant in terms of the relevance of the SERPS.
What you are now saying is that Google doesn't like sites that are optimized and sounds more like conpiracy theory than a real reason for not developing a PFI program, after all, weren't you the one that said, "At the end of the month, my sites could be listed 200th or worst if i didn't optimise them well and if i would have paid for them then...they would have "still" been listed 200th...you follow me".
So how would PFI affect relevance? In two-three months your site will be listed anyway, optimized or not. On top of that you weren't really asking for straight PFI, you wanted to see the bot nearly everyday and a complete reindexing every 2-3 days then ignored any thoughts that reindexing 3 billion pages every three days would pose any computational problems or place more demand on those "monster" computers.
Additionally, Google's current model seems to be working quite well, compare to Altavista which does offer PFI.
>>So here we go again, does it worth to lose millions because they don't want our highly optimized sites to hit the #1 spot too fast EVEN if our #1 site respect Google relevant rules?
No, I'm not going there again. You haven't offered up anything to indicate that they are losing any income, all you've done is proffer up an opportunity cost as the basis for changing their current economic model.
| 2:39 pm on Feb 11, 2003 (gmt 0)|
Google has PFI as has been pointed out before and it is called premium listings, want to guarantee your place then this is where you should be.
Webmasters are so desperate to rank high in the SERPS as Google is the best, they deliver the most accurate results and that is because they control the SERPS not a hundred commercial websites upping the bids every five minutes.
From a webmaster's point of view I see it as a challenge, I want to be at the top but only for results that would make the user happy, otherwise it is just a waste of bandwith, and that challenge is open to everyone which makes it even more exciting.
| 12:32 am on Feb 12, 2003 (gmt 0)|
Ok maybe i was not that clear.
I didn't change my mind regarding relevancy of sites listed #1 either by Google or by us with all the many modifications to our site since they will both pass through the same spam and relevancy rules.
This being said, it doesn't mean that Google share my point on this. Maybe for them most sites listed #1 by PFI are all cloaked and even if the cloaked page fits the relevancy filter, the rest of site that people will see may not be that relevant in the end. Just a theory here.
Chyio said that Google knows that PFI mostly benefits the ones who have the money and those sites have SEO backup which means that if for Google PFI=CLoaking then that could be the main reason for Google to not let us beat their system in just a few weeks? Well there's only one "Guy" here who knows ;-)
| 10:48 am on Feb 12, 2003 (gmt 0)|
I'd still like to hear your opinion of a reason pointed out earlier in this thread.
Imagine your pfi site would get crawled and PageRank for it would get calculated in 2 days. Like already pointed out, calculating your PR would mean calculating the PR of EVERY OTHER site in G's index. Now, after those 2 days, G would have recalculated the PR of every site in their index. What would they do? Would they apply the newly calculated PR to your site only, and apply the old PR to all their free listings?
Result: The search wouldn't be relevant anymore, and G would've wasted all the time and computing power used for calculating the PR of all the sites in their index. Why would they want to apply the newly calculated PR to the pfi sites only, if they've updated their whole index anyway? It'd be better to use the updated PR to all their sites, because this way they would keep their customers (= the people doing searches) happy, and because this way they wouldn't waste their resources in calculating PR that never gets in the results.
Result: G updates the whole index and does the Google dance in 2 days.
Result: This would mean that BOTH the pfi sites AND free sites would get newly calculated PR ANYWAY. So, only fool would PAY for what the other sites get for free.
Result: No one would pay for inclusion anymore because G updates its whole index in 2 days.
Result: G can start updating their index whenever they want because no one pays for inclusion anyway.
| 12:55 pm on Feb 12, 2003 (gmt 0)|
Well i don't have all the answers after all i'm the one who started this thread to receive some by you guys but we could say the same about Link pop for PFI programs at Inktomi, Fast, AV and the others my friend ;-)
Because i noticed that at Inktomi, if i add a page by PFI and on that page i have a link pointing to your site for example, well did you know that in the next update (in 24-48 hrs) your site WILL have one more link when you do a search for your site's link pop? So even if i'm helping others get more link pop, does that mean that everyone should close their PFI account?
Anyway, this means that it's possible to include many pages by PFI and recalculate the entire PFI and FREE db and that in a matter of hrs not days or weeks and this is why i don't believe the theory of GG don't have the ressources to calculate and recalculate all those sites link pop increasing not only your new PFI site PR but everyone else's PR also so there has to be something else...
No matter what we say here, i still believe that us, SEOs, will be given some sort of advantage by PFI and Google is not willing to give it to us no matter how millions they might be losing in the process.
Just my 2 cents,
| 1:08 pm on Feb 12, 2003 (gmt 0)|
It may loose some money in the short term but in the long term it gains the respect of users and this is why when you compare 1,000 webmasters nearly all of them will have 50 - 60% of clicks coming in from Google, PFI or no PFI.
Why change what is not broken.
| 1:22 pm on Feb 12, 2003 (gmt 0)|
>i still believe that us, SEOs, will be given some sort
>of advantage by PFI and Google is not willing to
>give it to us no matter how millions they might be
>losing in the process.
Wow, um, do you also subscribe to some idea of lil' black helicopters tucked away in a barn with a sliding roof down on the south 40 of the ol' google-plex?
What a conspiracy thought. Here's another possibility, though perhaps not nearly as much fun to dwell on.
Google decided to take a hike across the cyberterrain. During the hike, Google got some dieas, had some notions and was enjoying the view. Ultimately Google was on the rim of a canyon, having risen to the top of the canyon primarily by deciding aw ays back to NOT take the path many others chose, and to not look back with regrets on the decision.
Oh, so where are those other hikers today. They're hard to spot but some folks do still hutn for them. When one looks around carefully, they can be found, clutching the canyon walls, high above the river chasm, holding on for dear life on a narrow slippery trail.
So maybe, just maybe, Google execs don't sit around with a deck of cards going "red ten, ok, change algo parameter 14. Ha, 2 of clubs, that's one more month of no PFI for SEO's ... let them eat cake, muwhahahahahaha.
Sorry, but no PFI = intentional denial of income to SEO's doesn't sound like a part of the Google business plan, or anyone else's business plan forthat matter.
| 9:09 pm on Feb 12, 2003 (gmt 0)|
>Wow, um, do you also subscribe to some idea of lil' black helicopters tucked away in a barn >with a sliding roof down on >the south 40 of the ol' google-plex?
Of course not, but i wouldn't be suprised if we would find extra-terrestrials at the top of Google administration ;-)
Oops! Hold on, there's 2 men dressed in black knocking on my door now, i'll be back.
Huh... I have a telegram for you all...
Google is a the best Search Engine on the web today...stop
Google will not go PFI..stop
Thy shall not as seek as a SEO to go PFI...stop
Thy shall not cloak my page...stop
Resistance is futile...stop
| 3:26 am on Feb 14, 2003 (gmt 0)|
This thread gets funnier the more I read. You feeling okay, Mtlinfo?
| 9:29 am on Feb 14, 2003 (gmt 0)|
Well well well..isn't this Mr Google himself hehehe
At last we meet :-D Glad to know you found this thread funny!
You see, i searched and searched and searched a ton of your past msgs and many posts on this SEO board and many others on the web to find some kind of Press Release from Google on that specific subject but couldn't find a good answer so this is why i started this thread.
The difference my friend between you and JFK is that YOU...are still alive, so in order so stop this funny paranoia, you have the power to give us a final and definitive answer on that but the main question is...Will you?
YES...you are happy with your FREE relevant results and YES you are also proud of your PPC Adwords program but why leaving behind a lucrative PFI program that could bring you all those millions! Server incapabilities? PFI pages would make your serps less relevant?
In other words, why major SEs like Yahoo, AltaVista, Lycos, Hotbot, Theoma have their own PFI program and not Google?
If it's really a matter of relevancy then why have they all jumped in the PFI game? Are they all that stupid?
Can't wait fo you to close the debate here.
| 10:24 am on Feb 14, 2003 (gmt 0)|
> Can't wait fo you to close the debate here.
Mtlinfo, looking at the frequency of your posts, it looks very probable you will be the one closing this thread...
>In other words, why major SEs like Yahoo, AltaVista, Lycos, Hotbot, Theoma have their own PFI program and not Google?
>If it's really a matter of relevancy then why have they all jumped in the PFI game? Are they all that stupid?
Well, all these examples, combined, have only a fraction of Google reach. Their fall from prominence was caused by their bad (PFI and other) decisions.
So maybe, just maybe, they are all indeed that stupid :)
| 5:51 am on Feb 15, 2003 (gmt 0)|
"you have the power to give us a final and definitive answer on that" ... Well, no, I'm just some Googler--I can't read the future and promise with 100% certainty what Google will or won't do. Previously we've decided not to do pay-for-inclusion, and I don't expect that to change anytime soon.
| 8:19 am on Feb 15, 2003 (gmt 0)|
Thank you sir for giving us an answer on that.
>Previously we've decided not to do pay-for-inclusion, and I don't expect that to change >anytime soon.
...and of course i cannot push my luck and ask you why your firm took that decision...right? ;-)
| 3:15 pm on Feb 17, 2003 (gmt 0)|
|...and of course i cannot push my luck and ask you why your firm took that decision...right? ;-) |
If you read Sergey Brin and Larry Page's (the founders of Google) paper, The Anatomy of a Large-Scale Hypertextual Web Search Engine [www7.scu.edu.au] there is an Appendix about Advertising and Search Engines.
While this doesn't apply to pay per inclusion SPECIFICALLY you can tell from reading it that they feel that money taints the search results. Interesting to note that they said all of this then, and NOW they have full blown adwords and such on their SERPs.
But basically the bottom line is
|But we believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm. |
And there you have it, that is why they decided not to do PFI and why they originally didn't have ads at all.
| 7:55 pm on Feb 17, 2003 (gmt 0)|
75% of the market contrasted to 25%... Yeah, them Google folks obviously ain't got clue one about what brings the Consumers to a search engine 8^)
Poor ol' Google, sitting by the wall all alone at the PFI prom while the other SE's gather in the lobby, gossiping, complimenting each other, and just counting the days until they can pop the cork on a bottle of Mad Dawg 20/20 because they finally snared 25.0001% of the market.
Come to think of it, if Joe's Snowcone Shack carries 75% of the town's market without accepting kickbacks from the suppliers of Strawberry Pizazz Flavorings, Inc., even though every other snowcone stand pockets the proffered payment, maybe, just maybe, Joe knows that Strawberry Pizazz is not as desired by the consumer as Simply Strawberry, Joe's current best selling flavor.
| 2:24 pm on Feb 18, 2003 (gmt 0)|
kevinpate - totally agreed, PFI ruins search engines-- they should just try their best to make sure its possible for everyone to have an equal chance to be included (there will always be someone trying to hack the system so its never 100% fair) but the good news is that if you have a site with decent content and you submit to Google you'll probably be in their index in a good amount of time... and if your really desperate you can always do adwords. But I've always believed that pay for placement is a bad bad thing and that pay for inclusion is just one step below it.
| This 54 message thread spans 2 pages: < < 54 ( 1  ) |