Ruud Questions: Dave Harry aka the Gypsy

Ruud Hein

16 years ago

Stay Connected with Us!

I think I know Dave via Sphinn, somewhere... start of 2008? One of those people you see often.

Amidst a sea of corporate logo's, manga-style generated avatars and headshots, his avatar/profile pic stood out in the professional SEO community: sort of ripped-open shirt "here I'm sitting in my trailer with 12 lined up for the next hour" kind of look.

Odd.

Also noticable is that fame or no fame he was and is among a group of people who call a spade a spade. When tribes of people flock around a Sphinn submission (Ed.: am I getting my imagery mixed up? do tribes flock?) in "ooooh" and "aaaah", he's often one of those voices that says "well, sounds good but it's still crap".

He didn't seem to be one of those "new faces" that work their way in by tagging the word "expert" or "champion" to their name either: he throws his own solid stuff out there, posts that get young thinking and linking. So, I kept an eye on him and in turn have picked up series of interesting links and a bunch of good reads.

Emphasis and highlighting is mine.

One of your pet-peeves is SEO's throwing LSI (Latent Semantic Indexing) out there all the time. Only person I've read that's more about that is Dr. Garcia. If you had 5 minutes with the CEO of a company, apart from saying "it's crap", how would you explain to her what the problem is with their SEO selling them LSI-based services?

Sigh… great start… couldn’t let a fella even settle in a warm up could ya?

For starters, Dr. E has his own computational issues with LSI, mine tends to be more tempered these days as I have talked to IR peeps on both sides of that particular fence.
BUT there is a logic problem that keeps me ranting about so-called Google LSI based services; here’s the quick version -

The MAIN problem with Google/LSI is that there was this company named Applied Semantics, which Google picked up around ‘03, that had a semantic matching system. At the time Google felt the acquisition would, “enable Google to create new technologies that make online advertising more useful to users, publishers, and advertisers alike.” From the press release.

And there’s the first problem;it was for AdWords/AdSense. Used for matching the ads to the content - NOT for the regular (organic) search. In fact, there is NO EVIDENCE, that it ever found its way there. Unfortunately there are SEOs the world over selling services based on 'Google LSI' bla bla bla…and arse-holes pimp the BS.

Now why stop there.... strangely... that year they also brought in Sep Kamvar and the gang (of personalized PageRank fame) when they annexed Kaltix whom were working on a variety of things including, “developing personalized and context-sensitive search technologies” (press release) and Sep is now a technical lead in the personalized search dept. But it’s that ‘context-sensitive’ mention that we’re interested in…. moving along…

Early the next year, 2004, they ingested Anna Patterson and she worked on some similar topic detection with the beloved series on phrase based indexing and retrieval. If you spend some time with those, you’ll see not only potential semantic analysis, but a more vibrant probabilistic learning model that is better suited to organic search (IMHO). But even that doesn’t matter…

What matters most is that we DON’T KNOW and none should be selling such services. It’s perfectly fine to talk about semantics, even Yahoo had a patent out recently on the topic (Determining Semantically Related Terms) that is a legitimate discussion. It simply isn’t a service to be sold with the price docked on Google’s tail feathers.

Selling or preaching Google LSI is simply bad form… There are many potential paths to better organic matching. For more, one of many rants is here; Stay off the LSI bandwagon.

..huff… puff… lemme breath for a second here. Ok… let’s roll.

Along that line of thought... How is anyone outside of SEO to know what's real and what's not? Verifying SEO recommendations is very hard when people within the industry don't agree on basics like "only 1st link counts".

My goodness… I’m not so sure outsiders in any industry do more than make judgement calls as consumers…For me this thing of ours breaks into;

A. SEO really is an Art…
B. Professional SEO will still have a place

For the art, there is almost an intrinsic bond between the optimizer and their query spaces. I’ve been an analytics junky for years and scope by the control panels like keys on my piano and can feel when something is out of tune. Part of understanding what to believe and how to adapt programs comes from having that feel for what works and doesn’t. Like any art, that takes time and dedication.

This brings us to part B, better known as SEO is Dead. As long a search engineers are coming up with new and improved ways to index and regurgitate the world’s information, search engine optimization will be alive and well. This is why folks should consider paying more attention to the geekier side of life (we’re not wieners, we’re SOSGs). Not only will this help to future proof your SEO to a certain extent, you will also know what to test, what to look for and how to analyze the data.

In short, there is no ‘easy’ way. Consumers have conflicting evidence in all markets. It is the educate customer/client that usually gets the best experience. From you home renovator to the TV or auto repair shop… there are less trained and less scrupulous providers – ours is no different.

If links count as heavy as it is suspected they do, is on-page optimization still something we should be working on or is it usually "good enough"?

Well that all depends on if you believe links have evolved. We sort of ran by it in the opening tirade but one area that is old news, but new ground, is personalized PageRank. Both Google and Yahoo have a few patents on it and some of the related link text scoring gave the appearance, (potentially) of such things as LSI and behavioural targeting. If we believe this to be true (or a related form) – the optimizing internal link weights and link texts becomes important.

Regardless, link building is one of the most expensive (and risky) endeavours in SEO. You simply MUST nail the on-page as tight as possible to squeeze out every last drop. There are no lack of talented peeps I know that have made a tidy profit with rock solid on-site and some focussed link building. What people are best working out are strong concepts and themes throughout the site.

A recent toy lobbed out in patent from the folks at Yahoo shows an interesting weighting for on-page SEO. They talk about using the primary KWs in the prominent elements (Title, H1) and sprinkling other elements based on their perceived value in other areas of the page. It’s a simplified process… but speaks volumes as to what they feel about on-page SEO

You simply cannot afford to go light with the on-site…or the link building costs may bury you. One needs to bring ALL the weapons to bare IMHO.

Your friend says that with the economic times being what they are, he's going to prepare a bit of additional finances by "doing something on the web". Is he wasting his time? If not, how hard should he expect to work for how long to get anything close to even a minimum income? (average CAD $1300/month before taxes in Canada)

I’d tell them NOT to get some whacked out name for their blog/business that no one can pronounce, spell or remember. Straight away… that’s where we start. weeeeeee

After we had that out of the way I’d roll into the usual waffle about there being no free lunches and that it takes money to make money. It’s hard to say tho… after so many years of watching websites come and go and schemes fly up and down the flag pole. It’s like any business or job, its work. You work real hard and smart you can make far more… but that often requires hard work in the planning. Wasting you time would be to start without the motivation to see it through, or at least adapt on the fly.

I’d never discourage anyone, just like to be that bucket of cold water…

Having setup his site, affiliate and ad driven income, he now .... Yeah, what now? What should he do according to you?

Well start filling out the rest of the marketing tool box of course. Like any new born it needs a well balanced diet (of marketing goodiness) to prosper and grow. And if they don’t have the cash to hire some adventurer like m’self… then start learning and co-ordinating

Search engine optimization

Social Media Marketing

Analytics (and split testing)

Affiliate marketing (where appropriate)

Networking and promotion

Qualitative market research

…. And more. We haven’t even gotten into CRM, HRM and a bunch of other important acronyms!!

As you, and the gang here at SEP, likely understand…it’s the well balanced, efficiently planned attack that often wins the day. My advice at ALL times is for peeps to become well versed in all of the IM tactics and work the integration.

As a famous ‘Team’ member was fond of saying, “I love it when a plan comes together”.