SignalLinks Metadata – More Signal in our API
Late last week we announced that our newest analytic – SignalLinks – had left beta and was rolled out to all justSignal accounts. You can learn more about the SignalLinks analytic here.
Today we wanted to let you know that we’ve added the metadata for SignalLinks to our API.
We generate tons of metadata – or data about the Social Media content we collect – to facilitate our analytics and because we dig statistical analysis. Much of the metadata we generate never appears in our base analytics.
One of the key features of justSignal is our commitment to actually implementing an open data policy. Our open data policy isn’t a paragraph in our Terms of Service, or a vague statement on our site – it is something we consistently implement.
In accordance with that principal we publish every scrap of metadata we generate via our API services. SignalLinks is an excellent example. When we designed the metadata generation for SignalLinks we designed for two key views:
- Links – Volume data about specific links (how many times was it mentioned).
- Users – Who is mentioning which links, how often, etc.
The user view has not yet been released as an analytic. When we designed the SignalLinks analytic we realized that the Links view contained so much information that it didn’t make sense to try to cram in Users view – we thought it would be too much and, frankly, a little confusing.
But just because we aren’t using it right now doesn’t mean you shouldn’t be able to you.
The API release today includes both the Links and Users views – every scrap of metadata we are generating.
SignalLinks – Links View
The Links View contains two sets of information:
- The top (up to) 100 mentioned links
- Number of Mentions
- Date (month, day, year)
- Original URL (probably shortened)
- End URL (The actual destination URL)
- Domain (The domain the URL points to i.e., google.com)
- Links Summary – a summary of the data for all mentioned links.
- Date (month, day, year)
- Total number of mentions (content containing a link)
- Number of distinct links found.
- Mean number of mentions
- Median number of mentions
- Standard Deviation number of mentions
SignalLinks – User View
The User View contains the same two information sets, but organized around the user mentioning the link (not the link itself).
- The top (up to) 100 users posting links (by number of links posted)
- Number of times the user mentioned the link.
- Date (month, day, year)
- Original URL (probably shortened)
- End URL (The actual destination URL)
- Domain (The domain the URL points to i.e., google.com)
- User
- Links Summary – a summary of the data for all mentioned links.
- Date (month, day, year)
- Total number of mentions (content containing a link)
- Number of users who mentioned a link.
- Mean number of times a user mentioned a link.
- Median number of times a user mentioned a link.
- Standard Deviation number of times a user mentioned a link.
This is a powerful dataset for analyzing the links found in your Signal. Combining this with information you probably already have (metrics measuring the number of hits on specific landing pages, conversion rates on landing pages, etc) will give you a powerful tool to understand how much the appearance of those links on Social Media sites drives both traffic and conversion.
Social Media’s Big Problem – Marketers
From time to time our founder Brian Roy shares his observations about the state of Social Media, Social Media Analytics, tools and best practices (and any number of other topics) – and we share them here. These posts are Brian’s opinions, not the position of justSignal. That being said, Brian IS justSignal… so draw your own conclusions.
I hate to say it, but Social Media (and Twitter in particular) has a big problem… and that big problem is marketers.
I know, I know, marketers made Social Media – and setting aside weather or not that is true, let’s focus on the facts.
- Nearly any relatively popular topic is quickly overrun with marketers trying to get their message into your stream.
- The line between spam and marketing is non-exisitent in Social Media.
- Any analytical analysis of a topic is becoming more and more difficult as the topic gets filled with marketing.
- Traditional marketing approaches work in Social Media… get your message/link in front of enough eyeballs and some percentage will click.

Case in point. I own a Social Media solutions company – justSignal – and we have a Signal set up to track everything people say is “great” (via a variety of term searches and exclusions using our proprietary filtering mechanism) on Twitter. We’ve just released our SignalLinks Analytic in beta (you can learn more about SignalLinks here). So today I decided to look at SignalLinks for our everything great on Twitter Signal.
The results were disappointing to put it mildly. There is no authentic user voice in this data… only marketing and/or spam (you find the line there).
Here are the most mentioned links over the last 30 days:
|
Mentions |
URL |
Domain |
|
33511 |
http://www.tweeterspeed.com/ |
www.tweeterspeed.com |
|
12098 |
http://www.twtfast.info/ |
www.twtfast.info |
|
6818 |
http://tinify.net/5 |
tinify.net |
|
5641 |
http://twtexpress.info/ |
twtexpress.info |
|
5394 |
http://www.twitpwr.com/abuse.php |
www.twitpwr.com |
|
5032 |
http://twittfollow.com |
twittfollow.com |
|
3572 |
http://www.tweeterleaders.info |
www.tweeterleaders.info |
|
2904 |
http://followquick.info/ |
followquick.info |
|
2484 |
http://www.prankdial.com/fclicks/fclick.php?3 |
www.prankdial.com |
|
2212 |
http://WWW.TWEETERSPEEDY.COM |
WWW.TWEETERSPEEDY.COM |
|
2085 |
http://phobos.apple.com/WebObjects/MZStore.woa/wa/viewSoftware?id=304863000&mt=8 |
phobos.apple.com |
|
1632 |
www.twtmax.com |
|
|
1136 |
http://followersquick.info |
followersquick.info |
|
582 |
http://pollpigeon.com/what-did-you-think-of-adam-lamberts-ama-performance/t/79281/ |
pollpigeon.com |
|
349 |
http://pollpigeon.com/did-you-see-new-moonhow-did-you-like-it/t/79376/ |
pollpigeon.com |
|
346 |
http://4url.cc/1MM |
4url.cc |
|
306 |
http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewSoftware?id=305659257&mt=8 |
itunes.apple.com |
|
298 |
http://pollpigeon.com/do-you-like-emily-osment/t/75437/ |
pollpigeon.com |
|
248 |
http://boxcar.io |
boxcar.io |
|
160 |
http://www.nutritionaladvantageia.com/signup/ |
www.nutritionaladvantageia.com |
Let me make one thing perfectly clear – I belive that Social Media provides the best opportunity for opt-in targeted marketing. But when the Signal is so clogged with marketing and/or spam that adds zero value the only effect will be user apathy.
From a development/partner point of view, some of Twitter’s actions to “curate” seem rather annoying – but from an end user’s point of view they are doing exactly what they need to do. After all, at some point Twitter will launch their business model, and the two best bets are:
a) Targeted opt-in Ads
b) Analytics
Both of those revenue paths are put in serious jeopardy if users become apathetic because their Signal if full of marketing noise.
SignalLinks – Tracking the Link Buzz in your Signal
Today we’ve released – in beta, to a limited number of justSignal customers – our newest Signal Analytic, SignalLinks.
SignalLinks gives you detailed information about which links we find in your Signal, how often they are mentioned and which domains are mentioned the most. This information is provided for yesterday and the total for the previous 30 days.
Maybe most importantly we provide a graph of the number of link mentions versus the total number of unique links found.

This simple graph tells you the level of link “BUZZ” generated on any given day. The further apart the two lines get, the more likely it is that a small number of links are being mentioned repeatedly. In this example see 21-Nov:

The divergence of the lines tells you there was buzz around a small number of links being mentioned very often. The greater the separation the more “link buzz” there was.
Looking at the data we find that the link http://my.barackobama.com/tweetyoursenator-bitly/?source=101 was mentioned 1,808 times. The next closest link was mentioned less than 1/4 as much.
Link Reporting:
The report includes the most mentioned links for the previous day:

And the last 30 days:

Domain Reporting:
The report includes the most mentioned domains for the previous day:

And the last 30 days:

This information paired with our SignalMeter, SignalDensity and User Activity analytics provide you a comprehensive view of the activity for your Signal.
If you are not on the beta customer list for the SignalLink you should expect to start seeing SignalLinks delivered to your Inbox in the next 2 weeks. If you haven’t signed up for justSignal yet – get yours today.
As always if you have any questions just let us know.
Let’s Talk About Open
It is a simple philosophy really, you pay us to collect user generated content (or Social Media if you prefer) provide you analytics and engagement tools – you own the data.
Every company talks about open and data portability, we actually deliver.
How? We offer you two API levels of service that allow you to get all of the the data we collect – and the metadata we generate – out of justSignal. Maybe you want to feed it into your own Analytics solution… maybe you want to combine it with your marketing and sales data… maybe you just want to store it as a historical record. It doesn’t mater to us why, because it is your data.
So, what is actually in this data – because the devil is always in the details right?
There are two basic types of data you get from our Digest and Real Time API’s:
- Content
- Metadata
Content
The content section of the data is every element we collect from the underlying services. For example, for Twitter we pass along to you every single field we get from the Twitter Search API. It is all the content, every field. The same holds true for every service we collect from (Twitter, Blog Search, Backtype, Flickr, and YouTube).
When we add new services (more on that later) they will appear in the content section of the API.
Metadata
As we collect data from the services we examine the data for certain useful pieces of information. For example, we determine which of the terms in your filter were matched for that piece of content – which we use to generate your weekly SignalDensity report:

The following metadata types are currently included:
Content Term
The terms (or keywords) from your filter that were in the content.
Daily Volume Summary
The number of pieces of content we collected by service. Includes the total volume, number of unique users posting, average per user, median per user and standard deviation per user. For Example:
- Date: 2009-10-22
- Total Tweets: 685
- Number of Users: 565
- Average Tweets Per User: 1.21239
- Median Tweets Per User: 1
- Standard Deviation Tweets Per User: 0.89176
This information is generated for every service we collect from.
Daily Top 100 Users by Service
Every day we generate the top 100 users (by volume) for each service we collect from. Included in this data is the User information (user name and link to the user page if available) and number of pieces of content (tweets/blog posts/etc).
Tweet Links
We extract links from the Tweets we collect and generate information about that link. The information includes both the original link (likely shortened), the resolved link (what the shortened link actually points to), the domain of the resolved link, and a wealth of information about the HTTP transaction. For Example:
- Original URL: http://bit.ly/Fle3y
- Resolved URL: http://www.tweetmyjobs.com/jobpostings/2189294?src=1
- Domain: www.tweetmyjobs.com
- HTTP Transaction information including, HTTP Code, Transfer Time, Redirect Count, Content Type, etc
If you’d like a sample of the XML output of the API you can download one here.
As we continue to add services and create new and interesting sets of metadata they will appear in these APIs.
At justSignal we believe you can talk about open, or you can just do it. Our commitment to you is that your data is yours, you will always have the ability to extract it from our service and do with it what you will.
How’s that for open?
If you have any questions please let me know.

This year we are sponsoring SMAZ – Arizona’s premier Social Media event for businesses. The event is being held in 
So, let’s assume NBC has been monitoring – what is the upside now? How do you put the genie back in the bottle? 
When you consider that a few short months ago the prevailing meme was that 
Sometimes the best way to illustrate how justSignal can accelerate your strategy is by providing concrete examples of how others have accelerated theirs.

