Inktomi Optimization Frustration

Four Pages, Easy Terms, No Results

[RoosterTail]

Hey everyone. I have been studying SEO a while here on the forums, and I have been trying to get four of my pages into the MSN search results. The terms I am targetting are anywhere between 1 and 4 keywords in length, but since the site is scientifically oriented, the terms are not that competetive at all.

Back in late February, I signed up for a PositionTech account, and since then, I've been trult frusturated. I have four pages paid for and in the index, and according to PT, they've been refreshed on the 13th of March.

All in all, I'm just stumped. All of the pages validate, feature unique and dense titles, and META information.

I'm getting outranked in MSN as well as Inktomi's Pure Search left and right, and often only find my pages buried 100+ results down.

Whats the catch? My pages feature unique content, so I'm really lost here.

Any help would be great!

[RoosterTail]

No ideas out there? Or, is everyone just trying to keep their secrets - well, secret?! :)

Also, if it matters - I'm trying to concentrate these efforts ON MSN specifically, but want to do it without having to enter the Looksmart PPC arena. Thoughts?

[Shak]

Rooster,

sorry for not replying earlier, but I am amazed at your results.

Inktomi is an engine that quite easily will give us top 20 position with a holding page and corect metas :)

so I am lost to why this is happening, unless you have set off a filter or have too high keyword density.

I am probably not making sense, as I have NO answers.

Shak

[seth_wilde]

Have you been doing ongoing optimization? Have you tried emulating pages that are doing well?

[tennismaster]

Rooster,

I have posted about the mysteries of MSN ranking before,
I was in a similar postion a while ago.

You have obviously tried changed the titles and descriptions
and then waited for the 48 hr refresh, as I did.
Eventually though, after changing titles/description about
10 times, I did manage to get the page into the top page on MSN.
Some say links are taken into account by INK, but I have
not found any evidence of this.
My homepage has 500 links to it, yet is ranked below pages
with very few links.

Of course, if there are no `web directory` matches for your
keywords, you could join LOOKSMART and go straight to the top, but it will cost $0.15 a click.
I did this for a while until I got into top page via INK.

TM

[jimbeetle]

I'm with Shak on this, it's quite amazing. We've always done well on Ink for even somewhat competitive terms with just generally halfway well done pages. (Since Google doesn't like us we rely on MSN.)

I guess a couple of simple-stupid questions are in order. Are the pages indexable with text and text links and not images or javascript? What are the people above you in the SERPs doing? Do they rank higher for your specific search terms without apparent optimization?

>>feature unique and dense titles

As Shak mentioned, how dense is dense?

Other top-of-the-head stuff:

I've always had a 'feeling' that Ink likes older -- make that 'gracefully mature' -- pages. Not much you can do here.

Do links count on Ink? They had eons ago and I believe they might still. Do a check on alltheweb on some of the sites above you in the SERPs (link.all:www.domain.com -site:www.domain.com).

Other than that I'm at a loss,

Jim

[RoosterTail]

Thank you all for your responses.

I've tried to get as much information as possible, so here's a rundown.

In ATW, using the link search, and excluding my own domain - I have 5,242 search results to the domain.

In running an overall KW Density test on the pages over on Brett's SEW site, I'm seeing the following percentages:

Page 1: 3.14%
Page 2: 3.11%
Page 3: 2.45% (Ranks #1 in an Ink Pure Search, Nowhere on MSN)
Page 4: 2.88%

Clearly, that will need to change today. What I had not realized, was how much code bloat there was because of poor ASP scripting.

I've been told, on many occasions, that the on page content for Inktomi doesn't matter nearly as much as the META's do, so I really thought that I would be well off with a strong HEADing on the pages.

I will be adding in more keyword dense copy today on these pages. What percentages should I target? Should this be the only thing to change?

Thanks again everyone.

[jimbeetle]

I have 5,242 search results to the domain.

That just about knocks my belief that links carry substantial weight with Ink all to heck. If you can't get good placement on an obscure search term with 5000+ links then that theory doesn't seem to hold much weight.

Looking at some of my 1 and 2 results on MSN/INK for a slew of terms an eyeball test shows a strong title of 6 to 8 words and concise description of about 15 words works well with overall KWD using Brett's tool about 5%.

Keyword order does not appear to be as important as on Google. Same 1 and 2 rankings whether blue widgets this place or this place blue widgets. Go figure.

So, get to tweaking and let us know what happens.

Jim

[DaveAtIFG]

FWIW, very recent testing indicates that INK completely ignores keyword metas and alt tags.

[rmjvol]

...5,242 search results to the domain

Orphan pages?

RoosterTail, how many of your domain's other pages are in msn/ink? From other pages already in ink, are there links that spiders can follow from these other pages on your site to these 4 pages?

Without using your real domain, what's the format of your url for these pages?

rmjvol

[defanjos]

RoosterTail,

Maybe I misunderstood this statement:

Clearly, that will need to change today. What I had not realized, was how much code bloat there was because of poor ASP scripting.

but, ASP code cannot contribute to the page size because it is never seen by the spider/browser.

[skibum]

I'm not convinced links mean anything on INK, or if they do, very little. They just picked up some orphan pages from the dot-com daze and placed them near the top for the target keywords.

Don't think Alt tags mean anything, keyword metas are part of the Trusted Feed data required so there is probably some weight on those.

Generally I'd go with nothing but keywords in the title and description (just so its readable anyway) and text near the top of the page with keywords.

[RoosterTail]

rmjvol
The site uses ASP within it's structure, and contains querystrings. We have allowed some spiders to access these pages directly however, by eliminating the querystrings within the past month, and then implementing a footer nav of simple text links. ALl of these pages are 1 layer deep in the theme pyramid.

defanjos
I meant to say that the rendered/parsed ASP code is fairly bloated. It inserts comment tags for debugging, contains a lot of un-needed attributes, and many of the images are brought up through complete calls to the server, whereas relative links would be completely fine.

skibum
I completely agree about the links, or would have up until last week. This whole bout with Ink and MSN has thrown me for a loop, and I'm lost and confused. At my previous company's site, this same design and setup made a killing in MSN, and AOL back in the days they all came from Ink.

-- Thanks everyone! I've put the new text revisions into copywriting, and pending approval, we'll have that done. If that doesn't work, I don't know what will.

[Trodda]

I was doing some density analysis work prior to optimisation on Ink, and noticed one of my pages was perfectly set up to rank well, with the correct density, keywords in title, metas etc. Did a search on the keyphrase and found myself in only 13th position with a whole lot of rubbish above me.

Did a closer inspection and saw the top 3 ranked pages, had little or no density, no keywords in title or metas, YET DID HAVE have one or two of the following:
-the keyphrase either in the URL
-the keyphrase in the URL of an outgoing link
-the keyphrase in outgoing link text
-the keyphrase in the 'alt'tag.

Does Ink really put this much emphasis on this?

[minnapple]

Of course you need to see if your keyword is not buried by web directory sites and sponsored listings.
Otherwise this is a baseline start.

<head>
<title>three word phrase</title>
<meta name="keywords" content="three word phrase, three word, word phrase, three phrase, three, word, phrase">
<meta name="description" content="Compeling desciption">
</head>

<body>
<p><h1>three word phrase</h1>
good sales copy copy copy copy three word phrase . . . etc <b>three word phrase</b> copy copy etc.
<h2>three word phrase</h2></p>
</body>

[Trodda]

Thanks Minn. Didn't see any directory or sponsored listings above me.(was using Pure Search) Noticed that H1's and <B> seem to getting some sort of boost as well.

[Robski]

I agree with trodda's assessment. I've also tugged on my hair many a time trying to optomize for MSN with no real success. Things started to improve once I began to use key words and phrases in my url's. In my opinion this might be one of the most important factor.

[welshbabe]

Ummm i don't think that Ink spiders...you get what you pay for...just a thought

[Jaze]

Ummm i don't think that Ink spiders...you get what you pay for...just a thought

Yes it does spider, we have yet to pay for inclusion with Ink and we're in there. Also evidence in our log files of Ink crawling *.inktomisearch.com (if I remember the address rightly).

[nfinland]

"I'm getting outranked in MSN as well as Inktomi's Pure Search left and right, and often only find my pages buried 100+ results down."

I have the same "problem" as Rooster Tail. My site has a good ranking with several competed keywords in Google. My site also ranks good with the same keywords in Altavista and ask.com (Teoma also). The ranking on Inktomi is however way behind (+ ranking on Alltheweb has also dropped?).

Comparison:

Keyword 1

Google:4
Altavista:2
Ask.com:2
HotBot:28

Keyword 2

Google:10
Altavista:8
Ask.com:7
HotBot:81

I have looked at the "competing" sites and they have a good ranking both in Google and Inktomi. I´m trying to figure out what makes my site rank poor in Inktomi. My site is optimized with all the usual methods (title, description, H1, bold, keyword density, links text etc.). The problem should not be there? The link popularity is also quite the same for my competing sites. I´m thinking of every possible thing - my site is in php and I my navigation is an include etc. Taht should not make any difference?

Anyone having this kind of problems or some suggestions on what to try out?

p.s. The drop in rankings on Alltheweb happened quite recently. Could be off page factors, but I´m also trying to remember if I made some bigger changes on the pages that could have affected the rankings - poor memory and no documentation :-(

[gsx]

Jaze, you are right in that Ink does spider. It does follow pages, and updates the results approx every 21 days or so. This is the free service and many people get this without even realising.

The paid for option is different. It takes only the URL paid for and does not follow the links. You only get the page you paid for, one page. However, this is updated every 2 days on average allowing you to tweak to get their algorithm and your page into a harmonious match. Then you can tweak other pages, but you may be waiting a while before you get any results.

[Jaze]

gsx: I've also heard that as soon as you start paying, your free results tend to get dropped... not something I would like to test as we currently have about 30 pages in their index and damned if we're gonna pay for them! :)

[steveb]

I recently went from 15 to 501 (literally) on Ink. No clue as to why, beyond the dumbness of paying these guys for pfi in the first place.

[nfinland]

I have a couple of pages in paid incluison and about 60 indexed the "free". They haven´t dropped and I have ask about this before and the answer was that paid inclusion does not affect the other indexed pages.

[tigger]

I'll join the "Can't get my head around ink club"

I paid for a page a few weeks back and it's so buried I think I nearly found the last page on the net, I just can't get my head around it.

Anyone having any success with PFI pages and would like to share some tips

[pageoneresults]

Try matching your <h1> with your page title. We've been doing well in Ink with various industries. I've been using pretty much the same page formula that I've used since 1996.

<html>
<title>Blue Thingamajigs in California</title>
</html>

<body>
<h1>Blue Thingamajigs in California</h1>
<p>Somewhere in this first paragraph you'll mention Blue Thingamajigs in California again.</p>

Some <ol>s. Some <li>s. You know, standard page structure. ;)

[tigger]

Well that’s what I thought my page title & H1 are very nearly the same just a very slight change in the H1. I'll change it around to match the titles to see how I get on.

Thanks POR

[pageoneresults]

I should also point out that you may want to have a few links that read Blue Thingamajigs in California.

P.S. The title and <h1> don't have to be ordered the same. Just as long as the same keyword phrases appear. For example, your <title> might be Blue Thingamajigs in California and your <h> tag might read California - Blue Thingamajigs.

These are just a couple of the areas that need to be focused on. I've found that if the pages do well in Google, they typically do well elsewhere.

[tigger]

Cheers, I've made some changes so lets see what happens

[pageoneresults]

tigger, I'll look foward to seeing your reply here in the next couple of days. I'd be interested to know if my advice and the advice of others in this topic helped out. Ink is typically the easy one to get listings in. I can only make that claim based on my experience though.