Web Analytics Leadership, Social Media Leadership, and Correlations

Alright.  Here I am working through some analytics trending of my Twitter, Facebook, www.adronbhall.com, LinkedIn, and other various sources of traffic.  Here’s the frustrating part.  I don’t mean a little frustrating, I am talking about a-grade I almost want to break something frustrating.  I have bit.ly, Webtrends Analytics, a database with cross-correlated results, a Twitter monitor in another window, and other little apps to track what is going on.  Where does most of this data come together?  In my head.  Why do I have to manage everything this way, it is damn frustrating!

So here is an idea, and I am just thinking outside the box here (as I always do since nobody will let me in the box anyway :p ).  How about these things get merged into a good solid single interface.  Maybe even some of the environments get rolled together in addition to the tracking analytics and reports.  How about a solid dashboard that allows drilldown into the various mediums to allow actual interaction with the data and platforms?  Imagine a dashboard, that reports the analytics for all these things but then lets you just kick in and start tweeting or posting the latest bits to Facebook.  Why have a half dozen or more applications and web sites open.

What we need here is some technology leadership.

Others have discussed this recently and are looking for some results.  Ok, some of you readers may be thinking, “do you not work for Webtrends Adron?  Go build this stuff!”  Well maybe others and I am, but I am still frustrated at not having it right now.  It seems this is something that is slowly slipping by in this industry and we need a centralizing effort for all these data points to help in actually making sense of them.

A fellow analytics & marketing professional, Rodica Buzescu writes in her blog Morphing Through Time a post titled Web Analytics ? Leaving a Lot to Be Desired.  She lays out the current analytics situation very clearly.  There is no tool that really bridges the data together easily.  Some BI Pros out there may say “oh but I have all that stuff together”.  This may be true, but how much time and effort did that take?  What we need is a centralized management solution for this data.  Something to bridge the data together to make it actual, usable, and actionable information.

I have also had a number of conversations with Eric Peterson about various analytics on various platforms.  One such paltform is Twitalyzer (@twitalyzer), that he and @katzpdx have put together.  This is a prime example however, of a very useful tool, but something that lies outside of other tools.  The question is, how to integrate and what to integrate.

As an analytics advocate, practitioner, and developer myself I have one huge issue with a lot of my analytics that I use.  Webtrends Analytics & Data Mart help me bring most of those together.  However there are a number of things that are still disparate and segmented in the wrong areas.

So with this post as it stands, what do people want joined?  Where are all these data points that people want?  I think something good just may come out of this desire, but the ideas do have to come together solidly.  I have things I can?t disclose, but I know there are minds here at Webtrends thinking about bridging this gap right here and right now.

Brainstorming Into the Business Intelligence Clouds

Part II of my Business/Enterprise Intelligence Series, check out part 1.

In the previous part of this ongoing series of getting the business intelligence and analytics pieces of the enterprise intelligence world figure out, I hit on the first step of the entire process. Collection of available data points. Not the desired data points, nice to haves, or might haves, but the available data points.

In this entry I will step past that and get into a little more of the technical depth of how these data points will be connected in the enterprise systems. More to the point, how they are not connected and will need to be in the enterprise.

A Quick Review, The Systems

  • Point Of Sale (POS) w/ 300+ stores
  • Webtrends Analytics tracking the Awe Widgets Inc Website
  • Internal Accounting Software (IAS)
  • In-house Built Customer Lists for Sales w/ Excel & Access

With the four systems we have a huge amount of actionable data points. So how do we connect all of these data points? That is the thorny question that comes up. Again, just like during the discovery of what data is available, we have a simple first step for this process.

Figure out the origin and destination!

That is really it. Of course, as anyone would concede, there is a lot in between. Without knowing where we are starting and where we need to end up though, finding the in between is in vain. Sure, one can make the mistake as many business intelligence projects do, and start building architecture before anyone knows or actually can use the architecture. This is a severe mistake, and I myself have literally seen millions of dollars get wasted only to have to start a project over, right in the middle of the original project. Do NOT become one of those projects, find the origin and destination!

In this case I am again, going to use my creative side and determine the operations of this company. Awe Widgets Incorporated is currently using SQL Server and has in house skills developed among their staff. That provides an easy option of moving towards SQL Server Reporting Services (SSRS) and working with SQL Server Integration Services (SSIS) for the ETL bits. Whatever other platforms are in between will be easily connected with these two tool stacks. So now I know my multiple originations, and my single point destination.

Moving on to the fun bits.

Now we have to figure out how we’re going to get from our origination to our destination. This is where the trip becomes interesting. Just to make sure things stay clear, and list out what we have that we are working with.

   
Originations  
Excel & Access (Office 2007) *.mdb, *.xlsx, and *.csv/*.txt data stores
Internal Account Software (IAS) This one is a prospective can of worms.  Proprietary layouts, de-normalized & normalized data, and all sorts of redundant, non-atomic data.  This sounds like an accounting package right?  :p
Webtrends Analytics Data Exchange Web Services (DX) Webtrends web services provide REST style architecture, with the ability for data to be retrieved in XML, JSON, HTML, or other formats (we can add more if need be, just let us know).
Point of Sale System (POS) This system provides two daily exports, one at 6:00am and one at noon for processing.  The export format is *.csv.
   
Destination  
SSIS & SSRS SQL Server Integration Services used to connect SQL Server Reporting Services, with the core underlying data stored in SQL Server.

So now we are starting to get somewhere.  We now where we are, what we have, and where we want to go.  Time to wire some things up, so stay tuned.  That will be in the next entry.  Also, if you are planning on attending Webtrends Engage in New Orleans, but sure to look up the Enterprise Intelligence session that I will be presenting with Heather Crince of Webtrends and Tony G. of Orbitz?

If you missed the previous entry in this series, check out Where is the Other Data Tracking?!  Where are My Acronyms?!?!

Just Tidbits of Technical Tenaciousness

IE6/Windows XP

Microsoft wants users to give up IE6 & Windows XP.  I want users to give up IE6 & XP.  Please, please, please give those things up.  IE6 of course more than XP.  With what Windows 7 is though, go ahead, get on the bandwagon.  Win XP is moving the way of Win98/95 finally so let’s keep that movement going.

Free SEO Toolkit from ole’ Microsoft

So this looks interesting.  The thing is, as stated above, one has to kick the Windows XP Habit!  So drop it and check this out.  I am keen to see what anybody is coming up with as for these platforms these days.  Especially companies like Microsoft who have vast resources to put toward it.

Adbrite Switches to Webtrends

I see this regularly these days, as we are gaining tons of switchers, since our analytics generally rocks.  But this one really brings it home for me since I have worked with the Adbrite Team on the Implementation.  Great team at Adbrite, excellent implementation if I might say so myself.  If you are looking for a company to do advertisements on the web with, this is definitely a company to take a look at.

Encryption and Decryption Code Bits

I recently had to do some simple encryption and decryption for password storage.  I wanted to document this so a blog entry seemed in order.  So without any yammering here’s the simple class I created.

public static class EncryptionHelper
    {
        private const string CryptographyKey = "CryptKey";

        // The Initialization Vector for the DES encryption routine
        private static readonly byte[] iv =
            new byte[] { 220, 13, 41, 29, 1, 63, 73, 9 };

        /// <summary>
        /// Encrypts provided string parameter
        /// </summary>
        public static string Encrypt(string s)
        {
            if (string.IsNullOrEmpty(s)) return string.Empty;

            byte[] buffer = Encoding.ASCII.GetBytes(s);
            var des = new TripleDESCryptoServiceProvider();
            var md5 = new MD5CryptoServiceProvider();

            des.Key = md5.ComputeHash(Encoding.ASCII.GetBytes(CryptographyKey));
            des.IV = iv;

            string result = Convert.ToBase64String(
                des.CreateEncryptor().TransformFinalBlock(
                    buffer, 0, buffer.Length));

            return result;
        }

        /// <summary>
        /// Decrypts provided string parameter
        /// </summary>
        public static string Decrypt(string s)
        {
            if (string.IsNullOrEmpty(s)) return string.Empty;

            byte[] buffer = Convert.FromBase64String(s);
            var des = new TripleDESCryptoServiceProvider();
            var md5 = new MD5CryptoServiceProvider();

            des.Key = md5.ComputeHash(Encoding.ASCII.GetBytes(CryptographyKey));
            des.IV = iv;

            string result = Encoding.ASCII.GetString(
                des.CreateDecryptor().TransformFinalBlock(
                    buffer, 0, buffer.Length));

            return result;
        }
    }

.csharpcode, .csharpcode pre { font-size: small; color: black; font-family: consolas, “Courier New”, courier, monospace; background-color: #ffffff; /*white-space: pre;*/ } .csharpcode pre { margin: 0em; } .csharpcode .rem { color: #008000; } .csharpcode .kwrd { color: #0000ff; } .csharpcode .str { color: #006080; } .csharpcode .op { color: #0000c0; } .csharpcode .preproc { color: #cc6633; } .csharpcode .asp { background-color: #ffff00; } .csharpcode .html { color: #800000; } .csharpcode .attr { color: #ff0000; } .csharpcode .alt { background-color: #f4f4f4; width: 100%; margin: 0em; } .csharpcode .lnum { color: #606060; }

So simple enough.  Hope that is helpful to anyone interested in a super easy encrypt & decrypt example.  : )

Where Is The Other Data Tracking?! Where Are My Acronyms?!

Where is the business intelligence?

This blog entry may instigate just a bit.  It will also be a little long for a blog entry. You’ve been warned.  I suppose though, if you know me & the work I do, that is not really something new.  I see something wrong, broken, or otherwise and I am likely to point it out and describe it in detail.

As I roll into 2010 coding, implementing, and rocking with Webtrends, I have noticed something lacking in the analytics industry.  I will add the clause that obviously Webtrends has people thinking about these things and actively working on this topic, but what I want to point out is a general issue.  Where is the other data, where is the existing data?

It seems, even though some company’s kind of get to a certain point in connecting data points, not many really do.  The biggest reason is that most companies are just a few steps away from actually being able to do so.  The other even larger reason is, many do not realize what data should or should not be connected.

When someone starts pulling CRM (Customer Relationship Manager/Management), Analytics, POS (Point of Sale), ERP (Enterprise Resource Planning) data, and other sources into a single reporting repository we finally have real business intelligence.  Otherwise so many entities stumble through the land mines of data confusion.  I see this so much it really drives me crazy sometimes.

So how can a company or entity identify and connect these points of data?  It often starts with a ridiculously simple step.  At risk of oversimplifying things, let me just state the first step in getting out of the data confusion land mines is to first figure out your data.  Ask these things:

  • What data does the business have?
  • What data is currently used and available?

Do NOT ask what data you want, do NOT ask what may not be.  What you want to know first, and so many companies make this mistake, is to know what you know.  Do not, at the early stage of business intelligence information gathering start asking too many hypotheticals.  I promise the risk of failure increases exponentially for every hypothetical data point added.

Once you have identified what data is available, start figuring out how the data is related.  Once you understand the data you can then, and only then, make the huge leap to determining what data you want and how to get it to where you want.

Let me draw this out in a real world example.  Beware; I am using my creative mind now!

What we have so far, for Awe Widgets Incorporated, is several data points.

  • Point of Sale/POS Systems in 300+ stores.
  • Web Analytics (by Webtrends of course) tracking all sorts of great data points on the Awe Widgets Incorporated Website.
  • Internal Accounting Software (Almost ERP, not really)
  • In-house Built Customer Lists for Sales.

So there we go, four key pieces of tracking.  So how would they work together?  With a little further analysis (my key creative side now analyzes Awe Widgets Incorporated internal structure) and we find a few connections.

Correlation, POS to Webtrends Analytics

The POS System has a tracking identifier for customers which we can use to sync up with logged in users tracked via Webtrends Analytics.  This data can be used to derive who is and is not in stores purchasing.  In addition trending could follow the user flow to derive some actionable decisions on how to encourage online or store front shopping.  Just these two data points being connected add a lot of value.

Correlation, Internal Account Software ties to POS

Another data point tie in with the aforementioned POS & Webtrends data is the Internal Accounting Software (IAS).  The IAS holds information related to each sale, and other correlated information about how sales are going for the quarter, year, and other performance indicators.

Correlation, In-house Customer Lists for Sales

The sales department, in aggressive technical fashion has built a number of customer lists in Excel & Access.  The Access Application has a partially updated data store with a server based Excel file holding the updated piece of data about each of the sales person’s current sales.  I know, I hear it now, every developer that is familiar with this scenario screaming, “OMG, you have your data in Excel AND Access, and it is supposed to have integrity, and be aaaaaaaaaaaaaaaaggggggggggggggggghhhhhhhhhhh noooooo!”  But you know, this @#$% happens.  : )  When things are like this, solutions get creative.

Tying Together the Pieces

Alright, this is when the awesome nerd bits start to happen.  But I have covered enough for this entry.  In the following entries on this topic I will step through this first data finding mission and start discussions on how to connect these sources and get that data mart, warehouse, or other middle tier piece into action.   I will continue on and lead into how the data can finally start telling a real story.  Because in the end, the real story is, somebody needs actionable data to act upon.  Does it really matter where it is?

Check out Part II of this series.