comScore Total Universe Report
For the past three or four years, I'd look at Twitter's numbers on comScore, or Alexa, or Quantcast, or some other third party measurement service and I'd wonder "what would they be if they included mobile devices like phones and tablets"? Twitter was one of the first companies to have this issue as so much of its usage is on mobile. But now, there are many companies that have this issue. So looking at third party measurement data is becoming harder and harder.
Yesterday, comScore (a company I was on the board of for ten years but no longer am involved with) announced the Total Universe Report, "which provides audience measurement for 100 percent of a site’s traffic, including usage via mobile phones, apps, tablets and shared computers such as Internet cafes". Websites must run comScore's UDM tag to participate in the Total Universe Report. You can learn more about the UDM tag here.
I think this is a big deal. I'm eager to see the numbers for companies with large mobile user bases when the first numbers are out as part of comScore's April reporting (usually around the 10th of the following month). The internet has moved beyond the web onto mobile devices and that trend is only accelerating. I'm happy to see the measurement marketplace adapting and I am not surprised to see comScore leading the way.
We’re excited to bring this to market! Please contact us to learn more.
So do I have to add another analytics library to my mobile app? Couldn’t find much info about UDM without registering. :(I already have Flurry and Google Analytics libraries bundled with my apps. File size is already getting a little larger than I’d like.
I registered and did a performance review of their JS code snippet. It’s really awesome. They copied the Google Analytics snippet: the external JS script is inserted asynchronously and rather than calling JS functions commands are pushed into a queue (which is critical when working with async code to avoid race conditions). The beacon they send returns a 204 HTTP status code which means there is no response body (which is less work for the browser). I didn’t expect this level of quality in a first version snippet. Summary: The increase size to your HTML is small, but that’s not the important part. The biggest impediment to fast pages are external scripts because they block the page. This script is loaded asynchronously – huge win.
That counts more for mobile websites, not mobile apps.All my web calls in native code are asynchronous.Good info though. Thanks for looking into it.
Yes, another analytics library but we are open to partnerships with current mobile analytics firms for redirect opportunities much as we did with Omniture on web tagging. We’re for lower app footprints too! Direct message me if you have questions.
My understanding is that this does not include apps (which really limits how much I am ready to celebrate this)
Total traffic measurement is indeed a huge deal – as an API infrastructure provider (http://www.3scale.net) we measure API traffic (much of which is from mobile devices) across it’s surging across all clients – often becoming a significant fraction of their internet traffic. As time goes by straight web reporting just wont cut it.
API traffic is another issue and a really big onei’ve come to believe that most API traffic is machine to machine and not necessarily correlated to total audiencegetting a handle on that remains an enormous challenge
I think that would make sense intuitively, but it’s not really what we see across the APIs we manage. It’s true that B2B / SAAS etc. APIs tend to be more numerous, but B2C APIs (e.g. when used to drive a mobile app – which is happening more and more) tend to have much higher traffic volumes.So it’s more like “most APIs are (currently) Machine to Machine, most API traffic is Machine to User”We have customers who have both types of usage of their API (Business which tends to be M2M and Consumer) where the traffic in terms of calls on the B2C part is 10-20x the B2B part.
The analytics opportunity in APIs, at least from my DaaS point of view, is in a real-time dashboard. I’d love a Chartbeat for the Qwerly API.
Curious if there is opportunity to standardize around how APIs or web services to log when a user is viewing, interacting, or sharing content. Twitter posts does not mean audience measurements. The stream is always moving…
it is only going to get harder as machines become more peoplelike in their usage
Not necessarily, since it is a matter of placing smarts in engine where it is same artificial brain per C not matter the vehicle (2). Gets down to doing your artificial brain for the customer so that it moves from phone to tab to tv per being called on.Insofar as data that way, it would be the straight up per category/specificity as you remove human error. Then data will be data with a bigger pool delivering better results to use for the hunch in marketing plan.
i’m thinking more along the lines of the silent ip address – just devices
[…] and I’d wonder “what would they be if they included mobile devices like phones and tablets”? […]very good question. i thik that i would like to know the answer too! thnks for the post!
Definitely a big step in the right direction. This is just the first step in moving away from a “site mentality” in the metrics world. The notion of a site have non-porous site walls is going away entirely — making traditional metrics of limited use. Mobile, web tv, and tablets are just one aspect of this change.In just the past week we have seen the launch of news.me, Trove, and Bo.lt. This on top of the big adoption of FlipBoard, Zite, Pulse and the like. These next generation (some more than others) aggregation services point to a world where content is not always consumerd on the “site” from which it originated. Indeed, new infrastructure will enable content to route to where ever it is in demand — legally and with compensation flowing to content owners from those profiting from its use.This presents some new and unique challenges for a world built on “site metrics”. Google has made some inroads to solving this syndication issues with it canonical technology but much work remains for the Metrics industry to prepare for the new world where content “routes” by and between devices and sites. comScore’s approach is clearing heading in the right direction.Gregg FreishtatCEO, Vertical Acuity
Very interesting stuff here. It will be interesting to see how these metrics are defined. I wonder if a friends article on Facebook that I see but don’t open on my Flipboard counts toward the original articles numbers. If that’s the case views may be artificially inflated and the CTR will be driven down dramatically without any actual change in the habits of the actual site visitors.Properly accounting for the various APIs is going to give comScore something fun to play with for the next few years. Good luck to em.
I think the word you’re looking for is curation. It comes down to then who’s counted as the curator. In the end, that will be the consumer with the celebrated one being they who endow that power.Getting there will happen, but in the meantime, the delimna poses opportunity. The aggregation service that truly plays fair with the originator will win. It is just a matter of the originators supporting fairness over celebrity….and it is all a matter of money in the end.
I am not sure it is as big a deal as it seems, in that CPM based ad budgets are never moved by data. Audience data is a big factor in the purchase of media, but not the allocation of media budgets.I think the nascent Cost per Engagement (CPE) model is a bigger deal, if it can get on its legs and gain traction.
I wouldn’t be surprised if Google had access to some of this data in their own way. Doesn’t most web traffic go through Google one way or the other?
How so? Are you referring to usage on Android platforms, click throughs from search, or something else?
he has to be referring to google as the onramp to the web – and its reach through its applications and networks.
Exactly Mark. They have so many backways into all kinds of traffic. They could slice off Google Analytics data, mobile search data, geo, feedburner stuff, Ad tracking data, social, etc & extrapolate. My guess is they do it for internal intelligence purposes.
i catch Freds post through Igoogle!
You have to opt into that. Unless you are using an ad manager, I believe.
What I meant is Google has a lot of aggregated / blind data on us whether we opt or not. You’d be surprised how accurate extrapolations can get when you’re dealing with huge sample sizes.
yes, but not through where you think it is – probably yes to apis, yes toweb traffic and any of their own pixels, but beyond that, no. people aretoo nervous about the computers knowing us. It makes them feel less human
yeah google reports would be more reliable than comscore’s I guess
This is great, much better that what we had, but I think we need new metrics.As the web becomes more participatory and its limits less clear, I think that raw traffic data becomes less relevant. As an advertiser I don’t want to know that much about traffic but about interactions. Yes, for them you need the traffic, but I’m more interested in Twitter’s number of users, tweets per user and that kind of metrics.Anyway, what’s traffic anymore? If the app I use for Twitter in my phone is updating every 5 minutes but I’m sleeping, are those calls to the API relevant? It reminds me to those questions in internet usage polls that ask about time spent online.
unique visitors per month is still highly relevant
Why- couldn’t you make money out of highly engaged users that aren’t uniques?
My oh my…what does it matter? Getting past all the worker ants throwing around bs regarding data (much of it irrelevent), comes the real metric: store promotion leading to store visit (PC or Mobile), promotion leading directly to transaction (t is in store or over app) and so on.Then you are able to best get feedback on what you can do via up selling via networking bringing it more together.Keep pushing Fernando, for it takes guys with street smarts like yourself to push into the ‘not so creative’ hands info that is useful for their market.
When creating a website or a new venture, it is crucial to make sure it is accessible and easy to use for people on the go, depending on your content of course.However, not having a mobile version of your site is a missed opportunity. I wonder how many other sites get most of their traffic from mobile sources instead of computers?
Fred, I’m curious about your views on this latest brushfire about Google adn Apple collecting location data. Something about this post triggered the question for me.I thought most telecom and third party service companies do this kind of recording. At least, in Hong Kong, it’s always been well established that information on your phone is always used somewhere else by someone. Is this a bunch of hoo and bah? Or, do you think we’ll see any crackdown on the two companies for this?The web is entirely mobile, and that’s where we are going to be making our decisions. Shouldn’t companies have access to this location data, to help build a context for improved services?
i think the media loves to make privacy a big deal and scare everyonei’ve long accepted that lots of companies know everything about me and i’m ok with iti do my email on google’s servers. they could look at all of my investment activity and confidential info that flows through my email.i trust that they don’t because if they did, they’d lose their email users and that would be bad for them
so when you say they know “everything about you”, you mean in the machine sense, but no one at the company will actually look at your data?I mean, in principle, once you host with someone, you kind of trust them to not screw you over. But really, if they do give your info to third parties, how would you ever know it was them?
yes, i mean their machines know, their algorithms know, but their peopledon’t have access to it
That’s the problem with services like Diaspora. They try to take on facebook by making everything distributed. But they actually make everything less secure, not more (as they hope).Here’s why: when I am on facebook and I share with my 120 friends, I trust them not to share that information with the world. If it does, I know it was one of my 120 friends who I directly trusted — or facebook.However, when I use a distributed social network like Diaspora, and my 120 friends use various hosting companies like godaddy.com, foo.com, something.com, I now have to trust all the hosting companies. In fact, any hosting company that has 100,000 users actually has access to the personal data which is being shared by their 12,000,000 of their friends. That’s a big reach.It is not easy to solve this problem, but it happens when the hosting companies store your information on their servers, and you just retrieve it. I might trust my hosting provider, but why do I have to trust the hosting providers of all my friends?Diaspora makes my data less secure, not more. But it’s still very useful.Check out http://myownstream.com/blog…
Interesting replies. Wouldn’t you say, Fred that it comes down to: if you use a device to transfer something, that something can become public(?), if you want to transfer something completely private, send it in a pkg via land mail(?), just remembering that items inspiring the other party may share a tidbit with someone else via their phone/tab/PC and then that may become a little public.Otherwise, most of what is gained is numbers referring to a niche someone wants to buy data in relation to.
I’ve thought about your email in Google’s servers before. I’m also quite comfortable with my data being everywhere, but mostly because I think that there is so much info out there that nobody with access cares about mine.But I’m not sure I’d be that way if I managed highly sensitive information about important internet companies that compete with Google in many ways. I’m almost sure that what you say is correct, but the paranoid in me would be all the time jumping in my head 🙂
Truthfully, I have found that important data over time always leaks onto the net. I almost feel like it may be better to give it over in a slow drip, measured, in your own way, is a better solution.
Food for thought regarding server data collection:Imagine Bob is having a conversation with Jane. They’re sitting in Bob’s home and Jane is wearing a wire. The wire is not visible, it’s in her pocket, but if Bob asks Jane to reveal the recording device, she will happily.Does Jane have to make the recording device visible to Bob (ever, at first, all the time, never)? Is Jane’s behavior illegal, disingenuous, or on the level? Can she sell that data to other people without Bob’s explicit permission? Does she have to inform Bob directly or can she just tell him if he asks.Hopefully the metaphor is apt.
Actually, this has been an issue for them – they started an energy trading desk, (all those servers take up a lot of power, and they wanted to hedge power costs). Because they were google there was actual fear they could frontrun.
This is huge. So far we have been flying blind.
Very important step, though their press release needs a bit of Godin-esque “remark ability”Particularly for something so noteworthy.Thanks for sharing it. Critical as mobile Takes over.