A few weeks (or maybe months, it’s all a blur) ago I was put in contact with someone who was asking about getting into the analytics industry. I guess it’s an industry – or maybe a sector. I don’t know – but it sure as hell is something! Anyway, I didn’t know this guy, but I ended up corresponding via email and eventually met up with him to talk about how I got here, what I do now, and my general thoughts and opinions on the market/industry/sector/whatever. In retrospect, I wish I wrote more of that discussion down, and now I am. I’ll refer to him as Mr. X.

I never grew up planning on becoming an analytics – I didn’t even plan to be in IT at all. I always imagined myself somewhere in the hard sciences (astrophysics, high energy particles, or a member of the now crippled BPP). I actually tripped into the world of analytics by accident. I have been an IT contractor from about 1994 (yes I was young, but programming is just a long-winded kind of math, so it comes pretty naturally to me). It was a way to earn money (often under the table) while studied sciences in school, eventually going to university learn about aerospace engineering. I had some rough times in school (personal issues) and had to leave to take full time work. So I went back to what I knew best – IT. Eventually I worked at a company looking to get into analytics before it was even called analytics. They were the kind of company that valued performance over education, and I was quickly propelled into the role of Manager of Statistics. They created a department for analytics before Google even bought Urchin! So, they were pretty far ahead of the curve IMO (and it paid off VERY WELL for them to be).

Now that I’m here, though, I kind of like it. I still daydream about developing an interstellar ship engine, but some of the R&D I do now is pretty cool. One thing I stressed to Mr. X is that I had to work long and hard to get where I am today. Anyone can be a hack or fly-by-night “guru” in analytics. But to be a real contributor (and earner) in the industry, you can’t fake it. This is the most important message I can give to anyone thinking of getting into the business:

Like any other career, analytics requires a lot of time to be good at what you do. You can scrape a living out of doing odd jobs and throwing GA on a site, but if you don’t understand how it all works and why, you will have a pretty low ceiling in terms of earning potential. Most of my friends and family can attest to how little they’ve seen me over the past 15 years, but I can say with some degree of confidence that my skills and knowledge are amongst the highest grade in North America.

This isn’t to brag – and I said as much to Mr. X. I just see a lot of web programmers claiming to be “able to do analytics” because they can put a tag on a page – but analytics is just so much more than that. Any code monkey can implement a cookie-cutter solution, but you make the big bucks by knowing how to customize a solution. Customization (reliable customization that doesn’t need to be retrofitted every three weeks) requires knowing about how each business runs, what makes it successful, and how to make sense of the data that spews forth. Knowing JS just isn’t enough to be good.

I guess for me, analytics is personal. It’s important. It’s the future of online business, which is the future of business. I also think it’s the future of government, charity, and pretty much any organized group that handles money or numbers.

The other message I tried to impart on Mr. X was that analytics isn’t just “web analytics”. That’s a dominating theme now, but it’s not the only one. As stats manager, web analytics made up less than 30% of my work-week. The most profitable analytics my last employer invested in was numerical forensics. Membership lifecycles and monetization, financial forensics (which is a lot more than just balancing books), and working on cross-channel integration. It’s my prediction that the importance of “web” analytics will decrease exponentially over the next 10 years as the fad wears off, and people come to understand that it’s just another channel in there numerical analysis.

For the new people, cross-channel means integrating data from different sources together. For example: taking web traffic and mixing it (in intelligent ways, not randomly) with an inventory system. Or mixing a members database with support call-center data. It’s about taking data created and maintained in independent/compartmentalized silos and establishing relationships that provide information on which a business/organization can make decisions and take action on. It’s nothing new, by the way. Most major banks have been doing this with all of you for decades (how do you think the insurance branch knew you just bought a car and may need insurance)?

A lot of my friends (and many coworkers) consider me an elitist. The truth is, I am. I made sure to provide Mr. X with that disclaimer up front before we dived into our pints and 120 minute chat!

I don’t have any problems with people who want to play around in Google Analytics and look at how many people visited their WordPress blog last month. Cool – all the power to you. I think anyone who has a website needs to learn this stuff – at least at a very basic level. However there are a few (I think a lot, actually) people in the analytics space that treat the industry as a fly-by-night place to stop over for a few quick bucks. These are those ‘SEO specialists’ with one or two clients for a few months before falling off the face of the earth; or the snake-oil venders charging 40 hours to add a GA tag to your site then give you a quote for how much it will cost to enable custom page titles… We (real analysts) are a very rare breed, who are grossly outnumbered by the alternatives. The industry needs both, but my problem is presenting yourself as something that you aren’t. I made it clear, in no uncertain terms, that I wouldn’t continue to assist Mr. X if he ever intended to do that (he said he didn’t, and I think I may have hurt his feelings in suggesting it).

So that’s a bit about me, how I got into analytics, and my take on the industry (in a VERY short version of my talk with Mr. X). As always, if you have any questions at all, just post in the comments and I’ll definitely reply.

So Unilytics is hiring!The truth is that we’re always looking for top talent consultants (they weren’t actively looking when they hired me!) but we have specific growth needs.

If you know your way around Webtrends and SiteCatalyst, and you happen to be good at it, go check out the careers section of the Unilytics site. Note: we’re not looking for webdev guys who’ve played around with Google Analytics(we’ve already got a very talented GA specialist) – we’re looking for seasoned senior analytics specialists who know advanced tagging and understands the core principles behind reporting and can apply them on any reporting platform.

So if you are looking to work with a super strong team with A-list talent, get over to Unilytics’ site and apply!

More formatting issues. The new plug-in I got for synatax highlighting has some conflicting CSS with my theme… No time to fix right now, but the code is still there for you to copy/paste. Sorry!

So I’ve been back from Engage and while I don’t really have time to do this, I feel I owe some people an update/follow-up. In general, I think the conference was a great success for attendees (not so much for sponsors, but I won’t discuss that here). Where I think attendees did get let down was with my own workshop.

As I have been telling people, my workshop was supposed to be a hands-on problem solving session to help and educate simultaneously on the thinking and practices of asynchronous tagging; unfortunately a few people showed up thinking I was going to talk about Webtrends asynchronous tag.

So after I did my brief introduction (being given only 45 minutes is really not enough time to discuss things in any amount of detail!!),  I opened up the floor to field questions. I was expecting to get some questions asking me to show how someone may tag a specific asynchronous event or data point. Instead, the first question out of the gate was “why should I change my Webtrends tag for the new one?” I was a little confused at first – thinking that I misunderstood the question, but it turns out that one simple question resulted in a 20 minute derailment. So rather than getting 40 minutes to work through several attendee questions, I spent about 20 going over two short simple examples from what I’ve prepared on queue cards… [No  one in the room wanted wanted to raise any questions or ask for help with anything - of course there were several people willing to hang back after to ask me 1:1 questions!!!]

Anyway…. I think it’s safe to say that my session was a complete disaster and I feel like I let down the room. One person who stuck around to ask private questions requested that I put together a sample file for one of my queue-card scenarios. I haven’t heard back from this guy, but I figure I may as well get it up here so that I can drive any other Webtrends users back to this post who may be looking for more of the same. For the record, this situation is a simplified version of a much more complicated real-world issue I dealt with a while back.

The Scenario:

A client has a home-brew CMS that has many links on many pages (upwards of 100-300 links per page). Right now, they don’t have any control over the links in the CMS (this was some non-standard view-controller design whose abstraction made it managerially impossible to handle). The destinations of those anchors are not all unique, but the client has a need to track them to establish page real estate value and activity success. We could heat-map the page, but that won’t mean much if linked images are moved on the page. It also doesn’t simplify conversion tracking… So – we’re going to use some  asynchronous techniques to tag all the links.

One point I made during my intro is that I like to use existing JS libraries. What library you use really doesn’t matter, but my samples are in jQuery. I don’t advocate one library over another – I just advocate smart utilization of your time. Spend time making the important stuff in your code and leave the grunt work to others!

The Set-Up: XHTML

The XHTML to set up for this is pretty simple and straight-forward. It’s just a base container that calls a JavaScript file that will auto-generate a randomized table of links. The HTML looks like this (sorry for the poor formatting – the indentation was not preserved):

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

 
	Sample Dynamic Link Modification
 
	<script type="text/javascript"><!--mce:0--></script>
	<script src="jquery-1.5.1.min.js" type="text/javascript"><!--mce:1--></script>
 
	<script src="mod.js" type="text/javascript"><!--mce:2--></script>
 
 
This is a sample HTML file that demonstrates the dynamic linking event discussed during the workshop. The idea here is to go in and modify the elements after they are loaded on the page.
 
 
 
For the sake of this sample, we are using JavaScript to dynamically create a random list of links on the page, then calling a JS file titled <a href="/mod.js">mod.js</a> to find all the links after they're added and give them a class of "CHANGED". Note that the modifier did not change the link in this paragraph.
 
 
 
<button id="addClass" class="button">Add New Classes</button> 
Everything after this line is randomized!
 
 
 
 
<hr id="startline" />
 
 
	<script src="base.js" type="text/javascript"><!--mce:3--></script>

The Set-up: JavaScript

We are making a random table of links. This wasn’t the case for the real-world work, but it helps to organize everything in an easy to read way. I’ve randomized the size of the table to show that the solution is not hard coded. This uses the ready command, which is attached to ‘body’ not ‘document’ as you may see much more often.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

$('body').ready(startup()); //run as soon as the page is ready
 
function startup() {
	var newTable = makeTable();
	$('#startline').after(newTable); //inject the newly created table statement to the website
} //end startup function
 
function makeTable() {
	var statement = "<table width='100%' border='1' id='thetable' class='stuffs'>";
	var seed = 3 + Math.floor(Math.random() * 8);
	for(i = 0; i < seed; i++) {
		statement += "<tr id='row"
		 + (i + 1) 
		 + "'><td id='row"
		 + (i + 1) 
		 + "-col1'><p>Filler goes here.</p></td><td id='row"
		 + (i + 1) 
		 + "-col2'><p><a href='http://google.ca'>These links</a> would be going to different places in a live site from my example.</p></td></tr>";
	} //end loop
	statement += "</table>";
	return statement;
}

The Solution

The funny thing here is how short the code is for the solution vs. the set-up. Seasoned programmers won’t be surprised by this, but I like to point out just how simple the solution is for tagging. This is the crux of what I was hoping to get through to the room during my Webtrends Engage workshop, but I really don’t think I did a good job of it. Regardless, here’s how we go in AFTER the links are created to slap a class on there. Now, when you go back to multi-track your things, you’ll be able use the class attribute to name the anchor being clicked. This keeps your multitrack solution simple, and allows for almost any kind of naming system you want (note how I derived the name using a rule that pulls from existing element attributes from outside the element being modified.

1
2
3
4
5
6

$('.button').click(function() {
	$('#thetable').find('a').addClass(function(){
		var str = 'link_of_' + $(this).parents('td').attr('id');
		return str;
	});//end func
});

Below is a DRAFT of my summary blurb on the workshop I’m running in Webtrends’ Engage 2011  in San Francisco. Keeping in mind that it’s got to stay short, any feedback on how to improve wording or appeal?

Getting In Sync with Asynchronous Technologies – A Deep Dive On JavaScript Tracking

Patrick Mazzotta, Senior Web Analytics Consultant, Unilytics

During this hands-on technical workshop, Patrick will be jumping straight into the deep end of JavaScript tracking with a focus on asynchronous philosophies and techniques. Come prepared to learn about these principles in an interactive whiteboard session. You’ll leave with a deep understanding of the topic and resources to work with as you consider your tagging strategy for 2011.

UPDATE: it was an issue with the template. I’ve got this new one that looks great – it’s just going to cost me more time to post b/c I’ve got to think of/find relevant images to go with my posts now!

Let me start by saying I think Google Analytics provides a valuable service to a valid niche market. I’m not a Google hater, and I think they do some really great stuff…

So now you know I’m not prejudiced against Google, I can come out and say it:

Enough with the Google Analytics already!

These people who sing about how GA is all anyone ever needs, and there’s no reason to pay for analytics are driving me insane! I’m not going to make this a huge rant (I have other blogs for that!), so instead I’m going to just drop a few tidbits that will hopefully educate people in what the real limits are of Google, and why it really should only be used to serve a specific niche of all internet websites.

For the record, I’m not too well versed in Yahoo! Web Analytics yet, but I think there are enough substantial differences for me to recommend not extrapolating my Google Analytics comments to apply to Yahoo! Analytics as well.

So… what could possibly be limited in GA with such a super-power like Google behind it you ask? Plenty [seriously, I've half a mind to write a book on it], but lets focus on the three biggies:

1. You don’t own your data
   • Sure, you get to export some of the data, but it’s not technically yours. Google owns it, and uses it for a variety of internal purposes. Sure, you can trust Google not to give your info out to anyone else, but I think it’d be a stretch to assume they aren’t scraping the info for their own uses. Let’s be clear – you can opt-out of the sharing your data with others, but that only applies to sharing with other external members of that program (the name eludes me at the moment).
   • While you can export report data, you don’t get the raw log files generated from your pixel requests. Once aggregated, no one has that data but Google. You don’t have a chance of ever getting that raw data completely – meaning you could never reanalyze or re-utilize those logs for any other purpose. Namely, cross-channel integration (not just from a user-centric perspective, but even from a more holistic data integration perspective). I’m not the only one who wants to run audit comparisons of GA, Webtrends, and Omniture for all the same sessions… Am I?
2. Restrictions on how you track and collect data from your visitors.
   • This is easily the single most crippling point! You don’t get to know who a visitor is – you don’t even know what a visitor is. Next to new vs. returning, you’ve got nothing. Let’s list out a very short list of how this impairs analytics for advanced users:
     • No cross-domain tracking: yeah, you can try to use the same pixel on different sites, but then it’s all one big jumble of sites. There’s no ability to independently report on different internet properties with a master or integrated view.
     • No cross-visit tracking: you don’t get to string sessions together based on browser data. You don’t even get to report back identifiers from your website (like a user name or user id) to track highly mobile users likely to connect to your site from different locations. Doesn’t sound like a big deal? How many people do you know who use the same computer at the office that they do at home? But wait, there’s more! Campaign tracking across multiple sessions is now impossible too. I can hear some of you readers asking, ‘who could possibly care how a campaign three sessions ago affects conversions on the fourth session?’. Try any adult site; most dating sites; pretty much any site who takes more than one marketing contact point to convert… and guess what: there is evidence showing that daisy-chained campaign strategies are having more and more of an impact on online retail and subscription sites… So, maybe you should care and you just don’t know you need too yet.
     • No reliable profiling of visitors: I can ask any GA user how many of their outlier browser/platform combinations come from their most dedicated repeat visitors, and they’ll never be able to tell me.
   • Here are a few other profiling questions you can’t answer with GA because of their user identification restrictions:
     • How many family members (or unique logins) share the same computer to shop on my online store?
     • How many times as this visitor been exposed to the same ad campaign?
     • How has this user responded to last week’s campaign in relation to this one?
     • Is there a pattern to this user’s campaign reactions?
     • Do customers with highest volume of lifetime sales on my retail site have a favoured reaction/response to a specific [search engine/search keyphrase/campaign creative]?

There are going to be a few common responses to a lot of what I’ve pointed out, and I’m going to risk simplifying them into two main categories (yes, I’m list crazy today!):

1. People who didn’t know or consider the points. We’ll call them group 1.
2. People who don’t understand the value of the points. We’ll call them group 2.
3. People who don’t care one way or the other because Google is always the best at everything. We’ll interchangably use the terms ‘fan boys’ and ‘lost’ to discribe this last group.

Not accounting for the lost people, I expect to find the numbers of #1 and #2 to be nearly evenly split, with only a marginal lead by 2. If you fell into group 1, congrats. You’ve got some new info to empower you to be more informed in your analytics decision making. If you are part of group 2, grab that second cup of coffee and have a seat. There’s more to come.

Let’s take a breather for sec and think about my motivations here. I’m not going to sell readers on any products here (although if you want, the account managers over at Unilytics.com would be happy to take your call). I want to make sure people are doing what’s in their best interest, because that’s how I profit. See, I win when others win. It’s kind of how the service market works. The more value you see from your projects and initiatives, the more you’re willing to spend to take things to the next level. And that’s me. I’m the guy that takes you to the next level. Hell, I’ll consult for other consultants who get in over their heads.

See, GA is a great fit for some people. And they’ll pay for people like me to come in and help them up their GA game. That’s cool. I’m down with that. But some people need more that GA – and the consultants they get stuck with (for any number of reasons) don’t inform them of what web analytics is really about. IT’S NOT HITS AND VISITS – I’m seriously not going to debate that. Anyone who wants to challenge the point can argue with Peder over here. So these mislead organizations invest a few bucks with no real return. “I can see how many people bought stuff on my site!” Yay! Good for you! So can anyone looking at your inventory system…

So the point is: when you are in the business of providing professional advice on what organizations are supposed to be doing with their web analytics, you better understand the points I’ve mentiond above. I maintain that anyone who claims to understand the points and still claim membership to group 2 are either lying or aren’t in the business of making other businesses money.

There really are great uses for GA – I’m using GA on four of my sites right now (including this one). But it will be a cold day in hell before I recommend GA to a dating service, any organization with cross-domain monetization needs, or anyone in the gaming industry. Call me old fashioned, but I just don’t like screwing people over!

So there we have it. I know this has been a more aggressive post than usual, but I hope I’ve offered something in trade. If not clarification, maybe I’ve given you enough to start questioning what you are doing with your analytics (web or otherwise) and if you are doing what’s in the best interest of your organization in the mean time. I guess it’s a win for me if I have given you a decent perspective on why GA isn’t a good fit for all organizations. Maybe as a small bonus, I’ve empowered you to laugh a little and make a few snide remarks the next time you hear someone say, ‘Google Analytics is free, there’s no reason why anyone would need anything else.’

A side/corollary note: it’s important to me that my readers (however few of you there are) remember that I’m not a snob – I’m just a guy who believes in using the right tools for the right jobs. The future of analytics is one-to-one marketing, detailed user segmentation, and providing better, more customized and meaningful experiences for end users.

I updated my Expedia profile today so that instead of getting random emails about destination deals I don’t care about, they can focus on sending me just the routes that are of most interest to me (such as New Orleans, Vegas, or most sunny destinations). I don’t know what kind of segmentation they are doing with that, and how they’re correlating my consumption patterns with my preferences, but they sure as hell are missing out on massive up-sell capabilities on me if they aren’t doing something! Make my experience better – learn what I care about, and show me stuff I’m more likely to pay for. I don’t like having to work to give my money away, and most other people don’t either!

Let me know what you think of the two topics.I’m not going to post the full abstracts until I hear a decision from them on my application (which should be by December 15th).

EDIT: Since the deadline for submissions has past, and I’m waiting to hear back on the final list of selected topics, I think it’s okay to post the full abstracts which were submitted.

Where to Cross the Channel – Demystifying Cross-Channel Kick-Off Questions

Breakout session looking at two case studies on the pre-implementation phase of cross-channel reporting.

Come take a walk through two case studies where we will look at how they come to the decision to take their first steps in cross-channel reporting, and what those first steps were. If you or your organization are considering the plunge into integrated analytics but aren’t sure what your first steps should be, this session is for you. We’ll be focused on using case studies to look at how to ask the right questions to get your organization zeroed-in on the most important relationships and correlations for their needs.

• How does an organization decide when to take it’s first step to integrate?
• What kinds of indicators are there to determine which cross-channel approach is right for your organization?
• How to identify/choose points of intersection?
• How is the resulting data visualized and reported on?

This session will *not* cover technical integration techniques or how integration is done – that’s a much bigger topic! Rather, we will discuss organizational and planning needs of cross channel reporting, and walk through how companies from the two case studies both ended up with successful implementations without doing the same things.

Getting In Sync with Asynchronous Technologies – Tracking More Than Links and Movies

A highly technical workshop focused on using JavaScript to track dynamic sites and user events.

Interactive content isn’t just for Flash, and it’s not just tracking movies. There are as many asynchronous events on the web as there are ideas for ‘doing stuff’ – the reality is that there are volumes of useful information not being captured and not being reported every day.

This workshop is a technical discussion on the use of JavaScript to track things most organizations never knew was possible. We will look at questions such as:

• Where do we draw the line between useful and non-useful data?
• How do JavaScript frameworks affect site performance and coding efforts?
• How do you plan for scaling and growth?
• How do you collect data without writing interfacing code?
• Do you have to spend 40 hours adding a function call to every link on your page?

This is an open-format session with a focus on small footprints, fast code, and comprehensive tracking. If you are looking for a more technical session on tagging techniques and user event tracking, then this may be the session for you.

So, what is in this post? Basically, a semi-organized collection of my thoughts on what I’ve seen companies try to do and the pitfalls of trying to squeeze web analytics into what I call ‘old school’ organizational structure.

The article supports the following quote, and what started as a comment-based response has become this analytics post.

“In my view, integrating Web analytics with marketing is the future. Integrating Web analytics with IT (the SPSS/IBM route) is the past, and flawed.”

- Brian Clifton (c/o above link)

Let me drop my first disclaimer on you all right now: I could write a whole book on this topic, but if I had that kind of time I’d be playing more games with my friends! The scope of this post is narrow, and presents a highly biased (and what I’d consider insightful) perspective on a specific aspect of the topic…

So let’s get to it. I think this is an interesting perspective topic, but not because I see a whole lot of viable options. I find it interesting to see other organizations (I’m trying not to limit this discussion to only for-profit corporations) struggle with the idea of categorizing the role of analytics in their company. Starting with the basics, web analytics is not the only analytics. I think this is my single most dominating pet peevs with groups like the WAA (Web Analytics Association) that, while identifying it as such in their name, regularly consider their field of specialization as the only analytics out there. Data mining is analytics. Visualization is analytics. A lot of people will disagree with my position, but as far as I’m concerned, analytics is just a catchy keyword that sprung up over the past ten years as a direct derivative of analysis. [NOTE: Some of my colleagues would stipulate that it's all just a highly specified derivative of corporate statistics, but we're not going to touch that today!]

Why the rant on the analytics term? Because it’s critical to the following position:

The best case scenario for structuring an analytics group within a company/organization is for it to be it’s own independent entity (i.e. department, team, or whatever classifications you use in your organization). The head of this group should officially report equally to a CTO and COO or collectively to a board of directors.

- me

All fine and dandy, but that’s not really what Clifton was talking about, right? Kinda. Brian Clifton’s comments, while not direct, make very strong implications on the utilization and ownership of web analytics tools, which imply responsibility and steering capabilities. The real point to the Clifton quote (as I read it) is that analytics needs to be integrated equally with marketing and IT. The problem with this is that ‘Marketing’ will never take ownership of inventory databases, server maintenance, or client profiling data (think major Canadian banks that have thousands of terabytes of data on their client base nation-wide); and likewise, IT will never understand the driving forces behind marketing, or why option “B”, while technically more efficient, is not an equitable substitute for implementation option “A”.

Trying to group analytics staff under marketing will inevitably require that one of two things happen.

1. Marketers have to become more technically proficient (which I don’t consider realistic).
2. They’ll need to acquire dedicated technical resources under the marketing department that will lead to wasteful doubling of resources and skill-sets to other groups with similar needs.

I believe that by creating a self-directed entity within your organization, you make everyone (such as IT and Marketing) equitable clients of the analytics group. An empowered group lead [let's call this role 'stats dept. manager' or 'SDM' - the name really doesn't matter], will be able to achieve more and bypass circular processes by not being subordinate to any other groups – basically they can get more stuff done. Think I’m making this stuff up? Take a look around. The dominating players in the analytics arena today are banks, major financials, and bleeding-edge online media companies; all of which allocate independent groups to the role of analytics and data retrieval. These are the organizations paving the path with bleeding-edge analytical implementations and executions; and it’s not by accident.

These are a rare subspecies of humans whose backgrounds are equally technical and business savvy. Notice that I don’t say ‘business’ and not ‘marketing’ savvy. They only need to understand the value of marketing to business; they aren’t making or designing the marketing initiatives themselves. The SDM will staff his/her team with an elite selection of specialists with varying (but specific) skills. You can imagine them to be the SWAT equivalent of IT and business in the corporate world. In the most general case (every organization will vary slightly), you’ll get a data mining expert, a tagging and reporting ninja, and a cross-channel specialist. This 4-person team is a bare-minimum (the fourth being the SDM). This group of people will, under direction of the requirements compiled by the SDM, kick out just about any kind of analytical solution any organization could ever need.

I think an important point to make here is my anecdotal observations that encapsulating analytics under IT creates a common and well-known resistance to rapid change, adaptation, and access to resources. This won’t work for highly dynamic or fast-changing companies. However making them wholly answerable to marketing makes them an aggravation point for IT groups, and develops a tendency to cut corners for short-term gains with little regard to maintainability and longevity (in the scope of years or decades, which is were true analytics has the highest ROI in my opinion).

So why don’t all organizations make instinctively make independent groups such as this? Because that’s one hell of an expensive team, (small teams will cost between $200k-$400k per year, and teams of 6-8 can easily break $1million/year for best of breed). You may be able to get one or two roles overlapped in one person, but anyone who claims to be the best of breed in all four areas (data mining, report design & tagging, technology integration, and business-technical translations) is, quite frankly, full of crap. For example, I like to think I’m pretty damn good in all four areas, but I know there are better data miners and report designers than me. I can be a powerful resource that is good enough to get an organization off the ground, but I can’t be considered the end goal for any organization (fyi: this is as humble as I get people!).

So what do organizations do that can’t afford such a powerhouse team of professionals? OUTSOURCE! Call up companies like my current employer (Unilytics) and get all the power at a fraction of the cost. Sure, you don’t have full-time personnel, and you’re subject to their availability, but the trade-off is that you get A-list professionals at a fraction of the annual cost. Most small to medium sized businesses can’t utilize an entire full-time team of analytics professionals anyway, (and even some large sized companies need some hired guns to help out from time to time). Be smart, be savvey, and know where your boundaries are. If you can afford the team and you can utilize the team, nothing can substitute in-house A-list talent; but if you just need an bit of help nudging your organization to the next level, bring in the mercs to lend a hand. It’s what people like me are here for!

The caveat to this entire blog is that you can identify the pros from the charlatans. Web analytics is full of wanna-be hacks, and they are a dime-a-dozen. Being a guy in the industry for nearly a decade, I can spot them out from a mile away. Of course, organizations trying to hire into a new role will have a much more difficult time. If you’re a sadist in interviewing techniques, put your applicants in a room with your marketing lead, an IT lead, and big-picture operations figure (separately, never at the same time). Let them grill the candidates in whatever manner they see fit (in terms of how they understand the role). The person who gains universal acceptance is your pick. Don’t hire out of desperation, because replacing a hack is harder than you think (trust me on this). If your organization is willing to invest in an analytics group seriously, they can hold out for someone who knows what they’re doing. Is time is an issue, and you’re really not sure about how to hire? Then go to the mercs again. You’d be surprised to see how many of us (the pros, not the hacks) are willing to help you in the interview process. It won’t be cheap (you’ll be paying us by the hour, including research, prep, and review time), but you’ll definitely get your money’s worth. And if we can help you get someone who knows their stuff, it makes our lives easier when we have to help set them up!

Am I reaching? Overselling the concept? After all I’ve done and seen, I really don’t think so. But all the readers are free to make their own conclusions.

Thus the post… I’m looking to find categorize this site within Google Benchmarks, and I’m not quite sure where it fits. Suggestions and recommendations are greatly appreciated!

It starts with ‘Webtrends Enterprise Edition’. This is what I demo’d for my company as version 7; and what I will always argue was the regenisis of the Webtrends we know and :$sudo love today). It carried until some time unbeknownst to me when people started calling it ‘Webtrends Software Version’ or just ‘Webtrends Software’. I never got a memo, let alone an email notification or even a brand/logo patch! Nonetheless, ‘Software’ it is. I’m guessing it picked up sometime between v8.0a and v8.5. I think this was the silliest of the names, since, as far as I can recon, the On-Demand version is also software being executed in another place…

Now with the release of v9, we have a shiny new name. ‘Webtrends On Premises’, which I suppose is an accurate description, but is arguably (at least I’m arguing..) much more of a mouthful than ‘Enterprise’. If you want to change names around, be my guest. After all, if they own the intellectual property, they should be allowed to name their Webtrends software whatever they want; right?

I just can’t help but wonder how these kinds of decisions are made, and how much time and money went into the re-branding. At this level of operations, I don’t think gimmicky names is what’s going to tip the scales in the Webtrends vs. Omniture battle.