September 23, 2009

the Adobe Omniture marriage is a pipe dream

adobe-hqI and the web analytics Industry in general was flabbergasted by the announcement of the Adobe Omniture marriage and there’s been a big ‘huh?’ debate going on since; which I see no reason to elaborate on. Eric does (as usual) some good analysis/commentary in his two posts Thoughts on Adobe + Omniture and More color on Adobe + Omniture.

However; I would like to set free my long term belief about this, as I can see some positive momentum in justifying the deal – people getting lulled into thinking that it might make sense (it does not).

To put it politely, I am probably less optimistic about the value of this particular acquisition than most are (quite honestly, I believe it’s broken). Though, I am happy to see that we all agree about this not being a deal about revenue and/or bettering the overall margin of Adobe product sales, which is something Omniture cannot help with, quite the opposite actually. So it must be about the technology, and there is no doubt about the fact that uniting content and tracking is an inevitable next step for the analytics industry.

That said, in the desire to unite content and tracking (and the subsequent optimization) I assume that those involved (Adobe) understand that the current data-collection methodology – primarily Script/Pixel tagging – is a HACK. It was never intended to be more than an intermediate delivery mechanism and a way of circumventing the IT department until we got true access to the data stream. So the future should not, and is unlikely to, include a, as some put it “installing tags” task (something which we debated vigorously at X change earlier this month). If we think about the task (detailed event tracking) at hand for a second and extrapolate today’s web-page status into a future of web-applications (which is happening as we speak). In this scenario we must also accept the arrival of a Web OS of a kind – which could be the browser, which is what we are seeing today, but it could also be something like Chrome OS or a Adobe AIR permutation if you like (to keep a positive tone in this post). In this scenario, which I again believe is inevitable, we have some history to draw our conclusions from, which is, if I use MS Windows as an example, 25 years of event handling and logging. You would NEVER dream of tagging a windows application, why?, because the whole application is built around an event log which you can choose to tap into if needed. The Web OS (in whatever shape or form it arrives) will most certainly and without doubt include the same thinking.

In the above situation, buying a JavaScript tagging company (this might be too harsh a statement) based on yesterday’s idea is worthless; what they needed was actually to build this themselves. For adobe this is (must be) an inherent part of their future thinking, if they want to play in the Web OS game. Building it them selves could include acquiring a company for the engineers and analytics understanding; and for this, Omniture and their limited and by that calculation very expensive few hundred engineers doesn’t count.

If you want my quick quotable conclusion and a quick $100 bet; what we’ve seen is the best of Omniture and any succeeding years from here will be downhill in revenue and likely innovation, furthermore the content and tracking marriage will be won and or lost on a Web OS level – and thus likely to be between Google, Microsoft and perhaps even somebody like Mozilla.

Sorry for the rant, but being an Entrepreneur and as frugal as I am, it just hurts to see the right thinking (content+tracking strategy), but $1.8B wasted on the wrong execution.

Anywho, from one Entrepreneur to another: congrats to Josh, John and team. Well done!

(1) I’m told that SAP looked at Omniture as well (briefly though), which in the current situation (not thinking ahead) and without disrupting Omniture business would have made much more sense.

Cheers :-)
/ Dennis R. Mortensen (@dennismortensen)

Dennis R. Mortensen
25
  • Brock Landers

    You clearly have no idea how Omniture and Flash integrate their solutions. Javascript is not part of the picture for Video tracking. I feel bad for Yahoo.

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Brock,

    I think I have a very good idea (some might even call it knowledge) of Omniture and Adobe “integration” – I used to do nothing but compete with them, before we got acquired. And to set the record straight, I believe Omniture have some fantastic tech. BUT that’s beyond the point and perhaps I should have made my opinion even stronger. It’s the tagging exercise that’s broken, so whether you fiddle around with JavaScript, ActionScript or other platform specific tags does not change my opinion, about that, being wrong in the long term, however elegant it is done. I believe it is inevitable that we will end up tapping into an event-log from our Web-OS. (And I don’t believe this is a decade long vision either).

    Cheers
    d. :-)

  • http://makingmarketingactionable.com René Dechamps Otamendi

    Hi Dennis,

    Great post, thanks for the rant ;-)
    Imho, as you, I don’t think this is a good deal in the long term. Adding to what you said, another next step in the web analytics industry, besides uniting content with tracking, will be the management of that large amount of data integrated with the rest of the corporation’s data. We’re already seeing some large companies struggling with the amount of data and variables and combinations when they try to push WA data into their existing data warehouses. I agree that somebody like SAP looks like a better match as it would have allowed Omniture to integrate WA data with the rest of the data of the company.

    Side note: Maybe somebody will have the good idea to disrupt the scene and launch a software that would be priced by installation (as opposed of page views) and that would allow businesses to analyze internally ensuring ownership, security and confidentiality of data with a pricing that can be more easily forecasted (specially in these times of crisis)… Let’s be honest the current pricing model by pageviews used in the industry (besides the free tools ;-)) is not suited for everybody, specially when you start adding the options and professional services.

    Cheers from cloudy Madrid,

    René (@rdo)

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Rene,

    Good comment. AND I most certainly agree that behavioral data collection and data storage is becoming a commodity – one which you cannot charge for. I left that out of the rant – you know – to not go overboard :-)

    See you in DC?

    Cheers
    d.

  • http://www.predicta.net Andre Freitas

    Hi Dennis,
    Thanks for bringing my self-esteem back to me. I thought I was the only one thinking this was waste of great money (or abundancy of it!) and that I was the only one not seeing the benefits clearly.
    I also agree that WA will stop being a separate business/software and integrate into ALL user interaction interfaces. The industry will change and TAG is very old.
    Great insights. Finally somebody stepped up and said it.
    Thank you from Brazil !
    Andre

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Andre,

    Well put: “integrate into ALL user interaction interfaces” – and probably a more sellable version of WebOS (which for some sound a tad too futuristic)

    Cheers from NY
    d. :-)

  • http://www.rudishumpert.com Rudi Shumpert

    Dennis,

    This is a very interesting point of view. I am a developer by trade (read as code monkey), and I’ve done integration of both Omniture code & Google’s and even played with Yahoo’s code a little bit.

    Based on the ever increasing use of AJAX on sites for creating rich gui’s I really can’t see the JavaScript going away anytime soon, not even in the next 10 years. For me, part of the value of a non-free tool is not giving away the behavioral data of my users to the same company I pay for ads.

    I’m also a user of Adobe’s ColdFusion and I can see a real advantage, ie. making my job easier, to having a tighter integration of the 2 product suites. I am looking forward to this merger in hopes that more developers of not just public web sites, but high end on-demand web applications can better track visitor/customer usage. Not only to make more sales..but to really make a better end product/experience for the user. Sure, my job is to make improvements on the site that will facilitate conversions and ultimately sales via the web. My passion / enjoyment of what I do, is making the best site / application I can. If combining these companies together give me better tools to work with, I’m all for it!

    -Rudi

  • Jon

    So you’re saying that when I create a JPG that then goes viral on the web it will be Microsoft Windows that will give me the capability to study everywhere it went and how many times it was copied and viewed by uniques, not Photoshop or Omniture? Seems a little sci-fi, no?

    As for web pages being tag driven (HTML, CC, JS), I guess you could call the whole WEB a HACK? Maybe Microsoft or APPLE can give us superior(?) tagless web pages?

  • http://www.webanalytics.cl Andrés Flores

    Hi Dennis,

    I think the next 6-8 months will tell the complete story about this “happening”. Maybe Adobe is able to get something of all this, on the other hand I personally think that for Omniture’s current corporate customers the story is downhill in features yet maybe sunnier in terms of cost.

    I think is really interesting that you have taken this opportunity to move the topic to tagging. What are we talking here… proxy-sniffer technology similar to some alternatives discussed at the time tagging became mainstream and now is back in the picture mainly because of some solutions aimed to mobile device traffic measurement?

    Or something else?

    Cheers,

    Andrés

  • http://www.webworldtech.com/ Tad Reeves

    Well spoken! I’m glad someone finally pointed out that tag-based tracking is a total HACK!

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Thanks Tad.

    At least YOU and I are on the same page!
    (and we’ll work it from there.. he he)

    d. :-)

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Andres,

    I don’t think we’ve moved far enough ahead and to where we want to be, by deploying what you call “proxy-sniffer technology”. We (the WA industry) already have this in place and it almost entails the same pain and “tagging” tasks as a ordinary page script.

    Just as you can read e.g. the HTML DOM tree to get an understanding of the page, you will (I am most sure) be able to read a event log of what happened within your page (application).

    Cheers
    d. :-)

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Jon,

    >>So you’re saying that when I create a JPG that then goes viral on the web it will be Microsoft Windows that will give me the capability to study everywhere it went

    I am not advocating Windows as the future WebOS, I am however using it as an example of a well deployed Operating System (OS). – And when we move forward towards a webOS (such as perhaps ChromeOS) we can use Windows or any other OS for that matter, as a decent predictor for how some elements might play out. With this in mind the traditional OS event logging is prime to be replicated, and if thats the case, the need for manual tagging is partly removed.

    >>As for web pages being tag driven (HTML, CC, JS), I guess you could call the whole WEB a HACK? Maybe Microsoft or APPLE can give us superior(?) tagless web pages?

    Njahh. I Think HTML, CSS and JavaScript as you mention are splendid (or just OK perhaps) standards and I don’t see them as HACKs. What I do see as a hack is the way we (the WA industry) collect our data today.

    Thanks a lot for commenting

    Cheers
    d. :-)

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Rudi,

    10 years? thats a long time! :-)
    (I obviously, as you can read, think this will happen a lot sooner, not this year, not next, but soon)

    Cheers
    d.

  • http://www.rudishumpert.com Rudi Shumpert

    Perhaps 10 years is a bit much. If only we had the crystal ball to find out what would replace it.

    “I think there is a world market for maybe five computers.”
    Thomas Watson, (former) president of IBM

    Can’t recall the year ;)

  • http://www.omniture.com Brett Error

    Dennis,

    You are right. Tags are a TOTAL hack and completely uninteresting. They are just a means of collecting interaction and consumption data. But even at a higher level, collecting this data is also completely uninteresting.

    What I’m trying to do is increase yields for advertisers, improve customer experiences, and improve publishers’ ability to monetize their content. And that is where the data stream in question finally gets interesting!

    How I get it doesn’t matter that much. Tags are quite popular right now, but the data acquisition model is evolving. I’ve seen 10x growth in the last few years of both app-to-app data insertion (via my XML API) as well as bulk data uploads. I hope that in the future we’ll be able to tap into the “WebOS Event Log” as well. From a data collection perspective a decent analogy is to compare Omniture to syslog (http://en.wikipedia.org/wiki/Syslog) — but for marketing.

    So while I think that the data collection we do here is tactical… costly, painful, and completely uninteresting. What is important and interesting to me is that the data is collected centrally and value be extracted from it to not only provide decision support for business management, but also automatically take action on behalf of businesses and consumers. That is where the true value of Omniture lies.

    So, my advice: DOWN WITH TAGS! Well, not really. Tags still work well, so please, use whatever works for you to collect data. Just please collect and own your own data and make sure you are maximizing the value of this strategic resource in your org.

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Brett,

    Thank you very much for providing your viewpoint. Much appreciated!

    And we are perhaps not that far apart in thinking, when debating tags (or other data collection methodology). Speaking of the deal itself, your options/warrants/RSU’s might tell your heart a different story than mine :-)

    Cheers and all the best luck in integration (assuming this goes through, which I am most sure it does, unfortunately)

    dennis

  • http://www.omniture.com Brett Error

    That was purely my thoughts on tags.

    As for Adobe: The acquisition IS fraught with risk. As they love to teach in business school… 25% of acquisitions have a net positive effect. 25% are net neutral. And 50% are net negative. This one comes at enough of an angle which probably only diminishes the outlook.

    But, generally with higher risk comes greater returns. What if we do pull it off?! My equity has been dispositioned for much greater returns in the past than this transaction provides. At this point, I’m all in to see if we can blow that away together with Adobe.

    Thanks for your kind words. I look forward to connecting in person in the near future. Come to Adobe MAX! :)

  • Carl Rowlands

    A very thought-provoking post, Dennis, particularly in regards to the idea of an emerging web OS.

    With so many literate people reading this blog, I am a little nervous of venturing an opinion on what I think will happen, but I will try anyway.

    Certainly the idea of the web with an HTML codebase is set to decline further.

    However I don’t think that a single format is due to replace this. Sure there will be closed apps using Flex, but for every install of one of these, there will be a number of low-budget WordPress and Joomla-type systems; DIY in the sense of building a page in Notepad, but obviously a more rounded end result…

    Analytics should be able to link easily to these systems in the form of plug-ins. And in fact GA has plug-ins that connect seamlessly to WordPress and Joomla. These “bridges” would still need to be installed, even if a new web OS monoculture emerges.

    Taking what you say here as a starting point, maybe then the new “bridges” in a more reliable, OS-based system of tracking would involve integrating these content management systems with client-side and server-side processes.

    Wouldn’t it be tagging, but just not quite as we know it?

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Carl,

    Great input. Thanks.

    I actually believe we are very much on the same page, but that I (at least in thinking) have taken it one step further. When you envision a next step beyond HTML, which again, as you mention could indeed be Flex applications – I agree. When you mention that for every one of these there will be a dozen DIY applications (sites) that will use something else – I most certainly agree. There is no doubt that.

    But if we forget the exaggerate overuse of the word platform nowadays (something which used to be reserved for Operating Systems only), I envision and am quite sure that we will see a more common web platform emerge (I chose to call this the WebOS, but it might come under a different name). This platform (which I also suggest, might as well be the browser) will provide an underlying set functions such as an event log. So any application on top of this, would be stupid not to tap into the event log as it will become in inherent part of the platform and any applications (sites) developed on it.

    Think about this way. If you write a Windows application today, you could do that in any number of environments and any number of languages – but you would never develop a File System to store potential files, for that you would tap into the provided options from your platform (OS). I believe as in e.g. Windows, where event logging is integrated, we will see a set of more sophisticated base functions emerge.

    So when you say: “Wouldn’t it be tagging, but just not quite as we know it?” I would say no!, I actually think this will disappear. Thus my distrust in the value of the OMTR/ADBE marriage.

    Cheers
    d. :-)

  • http://www.tagman.com Paul Cook

    Hi Dennis,

    I agree the deal is completely puzzling – great news of WebTrends, Coremetrics and the Omniture founders I’m sure.

    I also agree that tags are a means to an end that have gotten out of control. Certainly the reason we started using tags back in the day was that cookie management in web-logs wasn’t good enough for deep marketing analysis. I’m not sure they’ve ever made life easier or solved the IT bottleneck.

    I’m not so sure a decent alternative is around the corner, hence why we’ve created a solution (TagMan) to resolve the issue around tagging so that IT can been removed out of the loop but still retain a degree of control.

    Paul

  • http://www.gartner.com Bill Gassman

    Great discussion Dennis. I could take my comments in many directions, but my overall reaction is one of disappointment that our industry is not evolving beyond proprietary purgatory. Early tech industries experienced quantum leap growth after standardization occurred. Remember the 80’s, with the battles for the physical layer (Ethernet vs. Token Ring vs. PBX) and the network layer (TCP/IP vs. DECnet vs. SNA vs. Wangnet vs. AppleTalk)? The big breakthrough happened in 1988 when a university professor proposed a “simple network management protocol” (SNMP RFC1067) to manage IP devices. The industry evolved from its proprietary ways to embracing open and interoperable standards in less than a year and whoosh – the Internet became possible. There are parallels here with the Web traffic analysis industry. Hundreds of vendors (if you count advertisers) each have their own tag and cookie, but there are no semantic or syntactic standards. Brett Error is correct; it is not about the collection, it is about improving yields for advertisers, improving customers experience and monetizing content. However, the impact on organizations would be much greater, and much quicker, if there was an open approach to instrumenting what we want to watch. Competition should be at the analysis and optimization level, not at the tag level. There is no sense in moving to a new approach to tagging, unless it is something open and interoperable.

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Bill,

    Thank you very much for your thoughtful commentary – which is (as expected) a much more eloquent conclusion of my “tagging” frustration. You say:

    “..my overall reaction is one of disappointment that our industry is not evolving beyond proprietary purgatory”

    Couldn’t agree more!

    >> There is no sense in moving to a new approach to tagging, unless it is something open and interoperable.

    Again – couldn’t agree more. This was a huge debate at X change in San Francisco earlier this month, where I for one at least, pitched the idea of not trying to go the route of a script wrapper – which is essentially yet another layer. I believe (right or wrong) that we won’t see an interoperable data collection methodology before the arrival of the “webOS”.

    Cheers
    d. :-)

  • Pingback: Visual Revenue | Using Microformats to extend Web Analytics tagging

  • Pingback: From Web Analytics & SEM To Business Intelligence