Category Archives: Uncategorized

Measuring PR

Since I am now writing in 2 places, I am testing which spot gets the most traffic and how many folks cross over to the Capstrat blog. I am wondering if my little blog generates more traffic than the Capstrat blog based on the same topic. So thanks for being part of my experiment. For those that haven’t bookmarked the Capstrat blog, here is the link to my latest post: Evaluating PR’s Return on Investment – Part 1

This particular post is part of a series I am working on that hopes to give some insight in how to go about measuring public relations efforts while also incorporating some of the tricks I’ve learned over the years of doing web analytics followed up with some potential case studies. If you have any suggestions please feel free to write.

A solution to backing out bad Omniture data…sort of

You’re in luck this week avid Diary readers. I am sitting about 30 feet from the Atlantic Ocean in the Outer Banks of North Carolina and its currently raining so I am inside writing. As an interesting aside, I found out that last week, Jimmy Buffet was renting the cottage I am. Would’ve been cool if I would have been here at the same time, I am sure he is a hell of a story teller. Probably doesn’t know too much about Web Analytics however.

I recently asked the masses to help me solve a problem with backing out data in Omniture. Unfortunately, the masses were silent which leads me to believe one of the following 1) no one cares 2) no one knows the answer. I am hoping #2 is the answer. So despite the silence, I’ve never stopped thinking about it and think I came up with the solution. Its not elegant but it works…sort of.

Since I am in the giving mode, I thought I’d share.

Step 1: You need to create an eVar to duplicate the value you capture via Purchase IDs. I called it Order Number.

Step 2: Create a SAINT classification for the newly created eVar Order Number. Make one of the columns ‘Status’ or whatever you want to call it.

Step 3: Export the SAINT classification for Order Number. Classify any order that was cancelled with something like ‘Cancel’

Step 4: Import SAINT classification

Now you have abilities to filter on things that are ‘Cancels’ or not. I’ve done it with some of our test data, and moved onto putting together a process to do this going forward. You can look at your Campaigns to see which ones are generating cancels. You can look at individual days and see how much was cancelled. I think the best use of it is actually with the Excel plugin as you could create data blocks with cancels (or without) and then use cell references to create cleaned up dashboards that would be filtered. Pretty neat stuff.

Pros of doing this:

It takes 5 min to put into action. Don’t need Omniture Engineering. Its a living table, so as things get cancelled or status changes you can change the SAINT file and doesn’t cause a ripple in the space-time contium

Cons of doing this:

Still doesn’t allow you to change the value of bad data (such as an incorrect dollar amount), just simply create a way to filter it out.

At the end of the day I still think Omniture needs a way to have Purchase IDs as a classification to easy move things around in a GUI like Campaign Manager as well as alter order data without having to go to Engineering. The idea is to make the tools as useful as possible and make it so your admin can actually adminster the data.

Omniture’s Developer Connection

As I’ve probably mentioned before, my gut feeling is Omniture eventually wants you to view them as the hub of all your web marketing (or even offline marketing) data. They’ve built things like Genesis to do drag and drop integrations with 3rd party data sources (DoubleClick, Responsys, etc). But what happens if the partners you do business don’t have a Genesis integration, or you don’t want to pay for the integration (as they are pricey)? Well, they recently answered that question with the introduction of the Omniture Developer Connection which allows you to create custom connections to data via XML, SOAP, Web Services, etc. They provide the toolkit, documentation, communal help, and the rest is up to you.

Since I don’t have any XML skills and don’t really have access to it internally either, I bought a bunch of books this winter to learn the basics with the grand plans of doing some of this stuff in the coming months. This is a great step in opening up the data and connecting all the dots how you see fit. Now I have a lot of reading to do.

Landing Page Optimization – Volume 1

As firms struggle with the current economic realities, advertising budgets are usually one of the first things to go. With less demand generation going on, invariably someone inside a company will bring up the novel concept that we should actually try to do a better job of converting the web traffic we do get and optimize the site to get better results. Most likely companies will start with landing pages, but it could cascade into other aspects of the site. Honestly, it shouldn’t take a budget cut to spur these efforts along, but for whatever reason that is often the catalyst. I am big proponent of testing things to find the optimal recipe for your customers and your company, but I actually would argue most folks are doing this incorrectly.

Part of why I think the methodology is flawed is the fact that I am not sure the tools can measure things the way you need to do a decent job of optimization. Most people are doing optimization strictly on what happens on that visit. So if the success criteria is conversions or click-throughs you’re measuring the tests against that goal. The problem though for some companies is that people aren’t going to do what you want them to do on that visit, but actually will do it the 2nd or 3rd time around. Without taking that behavior into account you might decide on a recipe that is limited to only what is happening on that initial visit. I would propose that to do this right things like ‘repeat visitors’ or ‘subsequent visits’ should be an important element in deciding the winning design for a page.

Example…we have landing pages for things like our ThinkPad x300 or x301. We drive demand generation to these pages but I think we’d be fooling ourselves if we thought people are going to come to the site and buy a $2,000 laptop on that visit. You can look at something like bounce rate for this page, which would tell you whether or not someone clicked on a link of the page to go further into the site and this might show a level of interest or engagement based on the landing page. In some cases, the visitor learns all they need to know on that visit and won’t need to go any further, so bounce rate might not tell you if they got any value from it.  I would say in cases like this, its more important to look at these visitors and whether or not they ever came back to your site at all, and associate that as the true value of the landing page.

I have some other thoughts on this and some of the limitations of current toolsets, but that will wait for a future post.

My Wishlist of improvements to Omniture (Vol 1)

Due to my recent blogging exile, I left a lot of things in queue. Hopefully over the next few weeks, I’ll whittle the queue down. Here is one I started in October…

With all the recent updates made to Google Analytics, it leads me to ponder what is Omniture going to do next to remain the dominant player in the web analytics space? I would say based on what I see from Google, the gap between the tools is actually getting a lot smaller, more so than the guys in Orem probably care to admit.

Google has a vested interest to make Google Analytics as robust as possible and give it away. Why you ask? Well, if you can prove the internet marketing you do such as Paid Search (via Google), banner ads (via DoubleClick/Google), or even things like GoogleTV are measurable than the likelihood of you as a marketer investing more money in these activities grows exponentially. Its genius. I used to think they wouldn’t devote a ton of money on developing it, but I see the light now, they are going to keep moving this thing along and really don’t have any limitation with regards to funding (obviously that is not entirely true, but they do have some cash) or analytics expertise (the company is run on analytics). So, in order for Omniture to keep its dominant position it needs to get better as well. What should they be working on? Here is one man’s opinion…

1) Campaign Interactions

I’ve ranted before about this but it bears repeating. The web analytics world is about to move away from the linear view of how campaigns work from an ROI perspective. So far no mass accepted vendor has cracked this code to show true influence of campaigns and the interaction effects of them on other campaigns. We are still stuck in the v1.0 world of evaluating campaigns based on the last click mentality. By doing so you potentially shut down upper funnel things that actually feed your ROI stars like Paid Search and Affiliate. To me this can be done reasonably simply by allowing the visitorID in the backend to actually keep a history of all tracking codes, pages, etc that a visitor id encounters, instead of client-side ways like Campaign Stacking.

2) Campaign Optimization

To do this piece you actually need item #1 to work. But the idea is if you understand all the interactive effects of campaigns on one another, you could then figure out the optimal mix to meet objectives based on spend and the lag effects. Say you have $5M to spend in a quarter, throw that into the planner and it would give suggestions for optimal spend.

My gut is Omniture would like to be the ‘hub’ of all your web marketing campaigns and reporting, and to do so you need to also have a way to optimize it based on results. To actually optimize smartly, you have to know how these tactics work together to create a conversion.

3) Genesis data integrations need more data

One thing that is severely lacking in order to make #2 work is the lack of the spend data coming from some of the Genesis integrations. Example, why doesn’t spend data from DoubleClick DART come thru the integration? How would you judge E/R or ROI without it? Seems like those kinds of things are non-brainers, but they are missing.

Another DART integration problem is the lack of things like ‘Creative Name’ aren’t passed. How in the world would you know which creatives drove the best results without it? The answer is you won’t. So why isn’t it part of the integration?

4) Easier way to back out bad data/Easier way to bring in data

The scariest thing about Omniture occurs when you all of a sudden get bad data (such as fraud, bad coding, etc). It is an ordeal trying to remove data from your report suites as it involves getting Engineering involved which means $. Omniture should absolutely have a way to filter out records like purchaseID’s, visitorIDs, etc to take the offending records out of the database so that your average admin can do this on their own. It would make the tool infinitely more useful to clean the data. Otherwise if you don’t feel like paying for the cleanup everytime, you’re stuck with the bad data which erodes the credibility and usability of the product. This to me is a no-brainer.

The same goes for bringing in data via Data Sources. It is friggin scary to do as it and a real pain to take the data back out if you make a mistake. Not sure why its so tough to architect a table that you can refresh if you need to instead of making the data permanent, but that’s the way it is set up today.

5) Allow for multi-suite Genesis integrations

So Lenovo is a worldwide company, doing a Genesis integration for something like DoubleClick at a country report suite level is cost prohibitive. At around $15k an integration, you’d blow out a serious amount of cash trying to do this in a bunch of countries. Why not do it at a global report suite, and also create a way to trickle down the reporting to the country report suites?

As a side note, I am actually experimenting with this idea. I have the DoubleClick integration set up on our WW report suite, and am thinking of dumping that data out and reimporting as a Data Source into the country report suites. Probably a way to do this with Omniture’s new developer toolkit (another post for another time).

6) Give each customer 5 to 10 hours a month of free consulting

Why would I advocate the free consulting? When budgets, etc get cut within a company, people take a look at the tools they pay for. Omniture can better help their customers understand the value the tool can give them by continuously helping them use the tool to solve their analytic questions. Prove the value to the customer, and they’ll use you for life. Make it difficult to use or charge them for auditing code, providing best practices, etc you  might lose the customer as they don’t see the true value because they can’t make it work right.

7) Greater access to Visitor ID level reporting

Omniture could open up the VisitorID level data so that clients can create better integrations with databases, etc to create profile information and understand things like Campaign Stacking, multi-session behaviors, and merchandising. I understand its a lot of data, and I am not advocating putting it in SiteCatalyst, but should be available in DataWarehouse or somewhere else. And free.

7) Use classifications across multiple variables via a master table

The biggest waste of time for me is the classification of data. It is maddening classifying all the tracking codes. Add on that a lot of times we capture the same tracking code 3 or 4 different ways for different expirations/attributions and the classification work needed to make the report suite run is painfully manual. Why not create a master table and have the user choose what elements to drag over for a classification?

Example: We capture tracking codes for campaign reporting, and also use that tracking code in Campaign Stacking. Why do I have to create another classification for Stacking if I’ve already defined it in Campaigns? Why not allow me to use elements of that loadsheet in a drop and drag model to classify it?

Same goes for things like Product Classification and Pages, why not have a master list of elements and drag over the ones you need for certain things?

8) Allow time segmentation across all reporting

Its odd to me that I can’t look at campaigns or product sales, etc by hours in the day. So we run a 4 hour sale on a particular SKU and I have no way to actually see what the lift is during that time in SiteCatalyst.

9) Extend the character limitation beyond 100.

I am sure there is a great database architecture reason that a lot of fields in the database are limited to 100 characters, but to me its crazy. A lot of times page names, urls, etc are going to be longer than 100 characters. As a result a lot of page names are truncated making the data unusable. Please extend to at least 255.

10) Make SearchCenter data rollup

We have a problem where the SearchCenter data is at a country report suite level to match the Google accounts in those countries. At the same I also want a Worldwide and Geo view of Search. But can’t do it.

11) Time-Stamping Segmentation

Discover has some great segmentation capabilities but one glaring hole is the inability to segment users using a timestamp. Example, say I wanted to create a segment of visitors that came from Campaign A, added something to a cart, on a particular date and watch their behavior over time. You couldn’t do it because ‘time’ doesn’t exist. I want a way to bucket visitors not only based on things they do, but when they do it.

and finally one of the most important to me…

12) Allow for more than 2 drilldowns in SiteCatalyst

Again because we are a global company, having global report suites isn’t the best option right now because of the limitation of only being able to drill down 2 levels. So say I am running a global campaign and I drill down by country, I am stuck to only one more level and won’t be able to drill completely down to the  media type or vendor generating the traffic or revenue. I want to go deeper. And yes, I know you can sort of do that in Discover, but Discover is also expensive and has limited accessibility depending on your licenses. I need to give the data to the masses.

and lastly, but probably most important for 99% of the universe…

13) Make the tool easier to use and understand for the uninitiated

Google’s new interface is basic and to the point. Things are labeled intuitively and laid out in a logic manner. People off the street can use it get the gist of it in 5 minutes. A lot of times I show people SiteCatalyst and they immediately freak out and get intimidated. I don’t have an easy answer what to do to solve for it, but its been obvious to me in looking at the 2 tools side by side.

So there you have it. My first volume of wishlist items to make the tool more usable. Anyone else out there have other ones?

Best of 2008: Music

Such is the time of the year for Best of lists…here is my music mix of 2008.

1 – Vampire Weekend  – Vampire Weekend (it took the whole year to grow on me, now I love it)

2 – The Black Keys – Attack & Release

3 – TV on the Radio – Dear Science

4 – Shearwater – Rook

5 – Fleet Foxes – Fleet Foxes

6 – Deerhunter – Microcastle

7 – Decemberists – Always the Bridesmaid Singles

8 – Bob Dylan – The Bootleg Series, Vol. 8

9 – Kanye West – 808s and Heartbreak

10 – Beck – Modern Guilt

11 – My Morning Jacket – Evil Urges

12 – Ryan Adams – Cardinology

13 – The Rosebuds – Life Like

14 – The Raconteurs – Consolers of the Lonely

15 – Kings of Leon – Only by the Night

16 – Helio Sequence – Keep Your Eyes Ahead

17 – Frightened Rabbit – The Midnight Organ Fight

18 – The Dodos – Visiter

19 – T.I. – Paper Trail

20 – Okkervil River – The Stand Ins

Dream Jobs

So…it seems like I’ve fallen behind again on the blog. It happens. This semester has been pretty tough, and maybe I’ll get into what I mean by that at a later date, but I’ve been thinking of other things lately as well.

Over the last few months I’ve had quite a few companies and headhunters contact me about various web analytics jobs around the country. Despite a crap economy, web analytics remains a hot job. In fact, in poor economic times this stuff probably becomes even more important to make sure you are actually do the right things with your web business/marketing. I’ve turned many of these down for common reasons like location or pay but it also got me thinking what kinds of jobs would I jump at? I’ve had no real plan to how I’ve gotten to where I’m at, wherever that may be, but the interest in me of late has got me thinking a bit more what is I want to do career-wise. A colleague thought a great blog post would be what would my dream jobs be. So without further adieu, my top 10 dream jobs…

1) Working at mlb.com in basically any capacity. Churbuck would say he’d want to be webmaster of redsox.com, but why limit yourself to a team? I say get the whole thing. I am a baseball freak so saying I am even remotely related to the Majors would fulfill a lifelong dream. MLB.com is a fantastic site with bleeding edge video capabilities and I think it would interesting to work there especially in web analytics and web marketing. Wonder if I could do that from Raleigh.

2) Working in the front office of one of the doormats of Major League Baseball in somewhat of a Bill James-esque fashion. Why do the Pittsburgh Pirates, Kansas City Royals, and Washington Nationals always finish in the cellar? Bad management and not developing their farm system. I’d love to work in the general manager’s office involved with personnel decisions, like drafting players, should we sign this free agent, etc. Tons of data analysis and using my years of accumulating useless baseball knowledge would be an amazing career. Again, I could brag that I’m in the Majors.

3) Google’s  Web Analytics Evangelist, Jr. – I’ve told him this before, but Avinash Kaushik has one of the coolest jobs in the world. I’d like to be the junior version of that. Of course the travel would probably kill me (recall the Japan trip?), not sure how he does that part of it, but its got to be fun teaching the virtues of actually measuring stuff to improve your web presence. Like bringing electricity to cave men.

4) Professor in Web Marketing at Appalachian State – I’ve always sort of thought about eventually getting a PhD and kicking it in the mountains of NC at my alma mater. Going to school any further would kind of suck, but I still think the job would rock. Get the summers off to travel or whatever. I think I could handle that.

5) Something in green energy. I firmly believe this will be like the Internet was, but probably bigger. One of the projects I am working on with my team at NC State is to create an electric car battery and bring it to market, so I’ve been immersed in this stuff over the last 2 months. This market is ready to explode as gas prices and climate change come to a head. I want to be a part of it to cash and be the Bill Gates of Green Energy.

6) Product Development for Analytics Software company like Omniture or Google Analytics. I love thinking about the kinds of problems marketers have in the web analytics arena, actually solving them for people by making improvements to the tools would be even better.

7) Brewmaster for Rogue. Best beer on planet. They could pay me in Double Dead Guy beer, or at least put it in my pension. Plus, I am fairly certain no one would care if I came to work with a big beard and wearing flannel everyday. What does this job have to do with web analytics or marketing? Absolutely nothing.

8) Salary cap space-filler/player in the NBA. You know there are guys on the roster of NBA teams that are solely there to give teams cap flexibility? Like I think Keith Van Horn makes like $4M a year just so he can be traded to make sure the salary cap isn’t exceeded in trades. Like salary cap buffer. I can be that guy. And I’ll do it for half that. And I’ll keep your stats while sitting on the bench. At halftime I can hand over regression analysis to the coach that says there is a negative correlation with points everytime the shooting guard touches the ball.

9) Video Game tester. I mean how awesome would that be? “Honey, how was your day at work?” “Well, I got stuck on that board with the ogres again, just couldn’t get past them. It was a rough day”. Also, working on the assumption I would wear sweatpants to work and drink beer at lunch without HR getting involved.

10) The Sports Guy’s intern. Those that don’t read Bill Simmons on ESPN’s page 2 probably have no idea what i am talking about. But I would get mad street cred within my crew which would be worth the pay cut.

If you guys out there have other suggestions, feel free to add…