Profile cover photo
Profile photo
Nicholas Martinez

Post has attachment
I'm pretty excited about this #VideoQualityReport  from

I rarely watch YouTube videos. They're a time suck and I avoid them at all costs. Except last night when I got stuck on a sub discussing great movie scenes. Short clips (~2-3 mins), but the video was slow. Since I was watching Monday Night Football, it wasn't a big deal for me to wait. 

But then +YouTube displayed a notification about slow streaming speeds and whether or not I wanted to read about the "Why" behind it. For want of a clip to watch, I clicked on it. 

This is a great marketing tool for #Google  to explain on a very granular level how the ISPs function in reaction to the service YouTube is trying to offer. YouTube explains how they parse video streams into smaller packets for more efficient streaming.

It provides a wonderful window into why a user might not be getting the best web all the time. It provides a wonderful graph to demonstrate how video streaming increases at night, and how all ISPs are not created equal. This is a great way to manage user expectations. It also empowers the user to make better consumer decisions.

Explore the link yourself to see how Google is trying to sell you on how you're not alone in your pain and how Google is working to assuage that anguish so you can get a higher quality product streamed to your device.
4 Photos - View album

Post has attachment
Discovered a new, curious thing with #GoogleImageSearch .

So I received a brandless screenshot of ABC Company's webpage ( pithily making pun of their webcopy. I wanted to see more of the website.

In lieu of searching the copy with " ", which I knew would yield results, I figured I'd search the image. I took the brandless screenshot emailed to me and cropped out the most recognizable graphic on it (the buses). I then image searched that (Example #2). It yielded no webpage results.

I resorted to a regular #GoogleSearch  with visible web copy, and found the ABC Company website.
I downloaded the original image and did a search (Example #1). This yielded 3 results.
I screenshot the image directly from the website and searched that (Example #3). That yielded 2 results.

This intrigues me since the image #searchalgorithm  isn't necessarily organizing the image merely by color & pattern recognition. If that were the case, then there should have been some ability of the  #GoogleAlgorithm  to locate the image on the web. 

The obvious low-hanging fruit is that an image's metadata is part of their organization process.

The curious thing about Example #2 (screenshot of a screenshot) is that Google can still recognize it as a bus but couldn't find its exact location on a domain. Example #3 (screenshot from website) was the kicker for me since it was still a screenshot (without the context provided by #1's possible metadata), yet yielded webpage results.

I will post about this later with a more popular photo and take an ad nausea approach to the copy of a copy of a copy aspect encountered here so I can see how much degradation there is between web searches.
6 Photos - View album

Post has attachment
I posted about Twitter being incentivized by positive speech. This is another example of their business model getting in the way of political transparency. I wouldn't be so conspiratorial if it weren't a removal of API privileges solely centered on these particular politically bent blog.

Their logic is sound: granting users an "undo" option via tweet deletion. But here, it's only politicians that are given the right to occlude what they have publicly stated. 

If a politician tweets something untoward, then they have the ability to make that gaffe disappear. Intriguing to me since Twitter opened up its firehose to Google but are peeling back the stream to a 3rd-party group trying to add value to the platform's ecosystem.

Post has attachment
As a #digitalmarketer  who's anti-advertising, I don't mind a world where I'm not being sold to. Alternatively, I appreciate of the fact that I'm receiving free content. I am aware of the free rider problem, but #adblockers  are simply a technological disrupter to the current business model, cited as "broken" in this article.

This article cites more technology "cleaning up" the online experience, while leveraging their place between a website & its sponsors and the end user. So, a new technology can cut another slice of revenue off the old technology's foundation. Nothing new there. If you're a believer in the free market, then you should applaud the ingenuity. 

What I see from this evolution is a shifting toward controlling the entire environment (i.e. Facebook, Twitter, Yahoo Gemini) for a user. Additionally, I see the growing importance of the user's traffic data, to make that ad dollar spend more "worth it." Lastly, there will be a curtailing of user privacy to create this better psychographic profile.

Pondering the complicity of nominally original content on unique domains paired with #linkattribution to bigger sites. Given the marketplace, a cogent, non-insightful website can still send links to a slew of targeted domains to boost domain authority.

The only issue, like any other, is that you use up the water in the well to send all links to one singular domain. You'd have to multiply efforts 100x to gain a modicum of #domainauthority increase on your target. Additionally, the single attribution to target domain becomes quickly apparent to #searchengines.

#Organic growth is chaotic and random by nature. Convergence of single domain purchaser, website templates, and template content layout curtail the domain owner from propagating efforts 100x in a way that might be beneficial to target domain (ideally). 

Averaging the aggregate sum is the dolt's way of "digging into the numbers."

Post has attachment
One of the things most intriguing about this move is that OF COURSE it makes sense to try to stymie negative speech online. But does it?

The profit motive attached to the authenticity of speech is what is so problematic here. I don't want some curmudgeon ceaselessly spewing his vitriol at me (bc he wants to), but I want to feel like I have the freedom to be grating and subversive online.

These social media platforms are incentivized to make sure my entire experience with them is positive, so they can structure their platform to prohibit negative affects being transmitted.

Walking through a packed city like NYC can be daunting given how the sheer volume of people can inhibit us seeing each other as human beings. It's easy to see all the people and just think of them as objects occupying your space instead of humans with histories and feelings and other loved ones they call their own. The veil of technology and separation from physical presence compound this difficulty to humanize the person on the other side of the screen.

I don't know.

Does #data have to tell a story by it's very nature? Does that make any #analyst a storyteller?

The words in a dictionary mean nothing unless they're sequenced and compiled correctly.

Some additional thoughts on the Einstein session of testing #Googleimagesearch :

- What constitutes "success" with this whole exercise?
Success is defined by pushing the limits of the search engine. Google is God. Given the breadth of it's organization of knowledge, anything I can do to confuse it constitutes success. Any discrepancy of original search results and the variable being tested means some change affected the search engine's capabilities
- What constitutes a photo? 
Facial recognition is the most prevalent attribute of a photo that we're used to be prompted about, e.g. Facebook's "Who's in this photo?" I removed Einstein's face so the engine couldn't simply place the image that way. Everything else seemed like logical connections to also test.
- What is photographic data?
A combination of colors oriented in a particular manner. I don't think I would've ever considered this if not for breaking the Google search engine functionality in various ways.
- What is data in general?
If Google is organizing all of this data (both lexical and graphical), and I'm trying to confuse it, then what constitutes data itself? I usually am testing the limits of what's identifiable (i.e. the "Googleability" of data) within a Google search. I'm still abstracting what this could possibly mean. My only presumption is that it's similar to the "photographic data" answer above: "a combination of identifiable patterns ordered in a particular manner".

Post has attachment
I was inspired to remove data in this particular testing of the #Googleimagesearch function. (This will be the end of using the #einstein image.)

I did a group select for removing the Einstein's face.
I did a group select for removing the color data from the chalkboard.
I then blocked off data in the image to cut out more chunks.
I finished off with the coup de grace of checkerboarding the entire image with deleted image data.

The #imagesearch engine seemed to have no problem with crops to original photos, so that's what inspired the removal of data from the image.

Conclusions (based on extrapolation):
- Data continuity seems to be part of the #searchengine . This is intriguing given what we know about functions (à la calculus class).
- The search engine mechanism must be searching for patterns on a very siloed level since it couldn't even closely place the checkerboard graphic.
- The search engine doesn't seem to be taking a holistic approach to analyzing the image. There was enough information in the checkboard graphic to place the it (I think), but the dark color values (remember: color is very important) threw off the results.
10 Photos - View album
Wait while more posts are being loaded