Have You Thought About This?

[Kurt]

As you may (or may not) know, Google recently launched Google Voice/Google Phone.

One of the features of this new service is the ability to translate a voice message into text and email it to you. This is a great feature.

But what does it really mean for SEO? It means Google has the technology to do the same to all audio and videos with audio and then can index the content of audio and apply SEO techniques, such as keyword density and LSI etc, to determine the rank and relevance of audio.

Not sure they are doing it right now, but it would seem to be a safe bet that they will be indexing audio in the future.

I predict that soon we'll be hearing about optimizing your podcast, mp3s and audio in your videos for Google ranking.

When you're writing the script for your next webinar, podcast or video, you may want to start weaving keywords into your presentation, just to get some practice.

Again, Google has proven they have the tech to do this already. Why wouldn't they apply the same technology to ranking all sorts of audio?

Who'll be the first to write an ebook about "audio seo"?

[Lucid]

The technology to convert spoken words into text has been around for a while. I bought a Covox card for my 286 twenty years ago. Granted, I had to train it first but it worked. Today, these things can recognize any voice without training. The phone company uses it.

As for Google actually using this technology to convert every file containing audio it encounters into text, that's another question. Some maybe, but not all.

Once in text format, indexing it would be no different than a regular HTML site. SEOing your audio would therefore follow the same rules, no need for a special ebook in my opinion.

[Kurt]

Originally Posted by Lucid

The technology to convert spoken words into text has been around for a while. I bought a Covox card for my 286 twenty years ago. Granted, I had to train it first but it worked. Today, these things can recognize any voice without training. The phone company uses it.

As for Google actually using this technology to convert every file containing audio it encounters into text, that's another question. Some maybe, but not all.

Once in text format, indexing it would be no different than a regular HTML site. SEOing your audio would therefore follow the same rules, no need for a special ebook in my opinion.

The ebook coment was a joke.

Here's the point...Of course speech to text has been around. But at what <cpu> expense? They are now willing to do it for FREE.

Granted, phone messages tend to be shorter. However, transcribing the audio in Youtube videos opens up a lot of doors .

I've never heard another person even mention the concept of optimizing audio for SEO and believe it's a concept that should be discussed by those that what to stay on the cutting edge of SEO...Not to mention the fact that Google is likely storing the text it converts from these phone messages to profile users in one way or another.

[Joe118]

They're not doing it for free. Your voice is a training sample for their algorithm. YOU are helping training THEIR algorithm for free

Google needs a large sample of regular conversation, to create such things as spoken-word-frequency maps across e.g. locales, age groups, and many other demographics. And as you said to profile user interests, and for many other as-yet-unknown purposes.

[Steven Carl Kelly]

I wouldn't mention LSI, Kurt.

[Oling]

I have been haring about Google voice for the last few weeks but as far as I know it is still only available in the US. Is it available now in Canada?

Thanks
Oling

[Kurt]

Originally Posted by Steven Carl Kelly

I wouldn't mention LSI, Kurt.

You just did, Steve.

BTW, I'm not a big believer in LSI for SERPs. Google bought the tech to use with Adsense, not SERPs. But the concept of related words is one that should still be discussed and understood.

I believe the concept of LSI can also be applied to voice and conversation particularly if they want to include relevant Adsense type (contextual) ads to pages containing videos.

Maybe Google will use speech to text indexing simply to target better ads for Youtube? Which in turn could mean the same for Adsense on pages with Youtube videos imbedded.

I don't think we should discount LSI for this purpose (and many others), just because you wouldn't mention it.

[Lucid]

> The ebook coment was a joke.

I know.

> Here's the point...Of course speech to text has been around. But at what <cpu>
> expense? They are now willing to do it for FREE.

Lots of companies do things for free. This is just another "hey, look what cool thing we can do". Maybe they are doing it for branding. On the other hand, they are building an operating system. Microsoft bundled a lot of utilities like Internet Explorer as part of Windows. This speech-to-text software could be a bundle inside Google OS.

In the end however, using Google's free software we will be exposed to advertising for the privilege of using it for free.

Joe118 is probably right too. Excellent way to test and improve software.

[Steven Carl Kelly]

Originally Posted by Kurt

You just did, Steve.

Yeah, but the more time I've spent looking at LSI, the less I feel that it bears much effort.

Originally Posted by Kurt

BTW, I'm not a big believer in LSI for SERPs. Google bought the tech to use with Adsense, not SERPs. But the concept of related words is one that should still be discussed and understood.

If you have about 15 minutes, take a look at this video:

Warning - “Advanced” SEO Technique DOES NOT WORK | StomperNet

[Kurt]

Originally Posted by Steven Carl Kelly

Yeah, but the more time I've spent looking at LSI, the less I feel that it bears much effort.



If you have about 15 minutes, take a look at this video:

Warning - “Advanced” SEO Technique DOES NOT WORK | StomperNet

Thanks...For some reason, I could only watch the first half and the video wouldn't download any more...I'm using a mobile, "cell" modem, which is likely the problem.

To be honest, I didn't learn anything in the first half of the video. I wrote in my own private forum years ago that I didn't believe LSI was a factor in the SERPs.

I have also posted about the weakness of LSI and how it needs to constantly change. 10 months ago, the words "president" and "obama" would not have been LSI related. They are now.

I also wrote in this forum years ago about text vectors, which is basically the same thing as described in the video.

And they didn't address the point that Google bought the company that had developed LSI, so the patent wouldn't have been filed in Google's name. But like I said above, I believe that they intended to use LSI for displaying Adsense, not their SERPs. Maybe this was addressed later in the video?

I still believe the "concept" of LSI as it pertains to related words is important, and have for my 13 years or so of SEO. And it really has little to do with the backend tech used by the engines.

If we accept the fact that aboout 40-60% of all search queries performed are "unique, one of a kind" search phrases that are impossible to really optimize for, because they are so unique. Google and all engines struggle to find relevant results for these type phrases.

And, we accept that every word on a web page is a "keyword" (except stop words), then I feel there's only one logical conclusiion:

We want to have as many related keywords on a page as possible, so they form as many unique combos as possible.

For example, if one has a page about rottweilers, using what's known as "LSI" related keywords, such as rotty, dog training, potty training, leash, collar, etc.

It isn't LSI at work, but the result are the same. It lets pages that use many related keywords have the best chance at the 40-60% of the traffic created from one of a kind searches, such as:

"how to potty train my rotty using a dog collar"

I doubt anyone's optimized for that phrase, and I doubt it will be searched for more than once (without the bias of someone trying to prove me wrong). But these types of search make up a major percentage of search traffic.

While I do keep up with SEO trends and info, my basic "on page" SEO strategies haven't changed in 13 years and most SEO gooroos make it too complex. It's basically a numbers game:

Put as many words on as many pages as possible, and link to those pages.

Get this part down, and the optimization will be much easier.