Web Analytics


…a curious person is one who explores first and then considers whether or not they want to accept the ramifications — Seth Godin

Web Analytics

Help me bring water to 15,000 people

UPDATE January 1, 2010: thanks everyone for contributing. Together we raised $2591 for UNICEF.

I’m a big fan of Unicef and the work that they do around the world. Every year they are one of the charities that our family donates to. Normally we try to do something that can really have an impact, like helping build a school, improve literacy or health care options for others.

I have been fortunate this year to have had the support of many from the UX community and would like to “pay it forward” with a little side project and raise $20,000 which can purchase a water purification unit used to provide drinking water for about 15,000 residents or displaced populations in cases of emergency.

While that sounds like a lot of money, it really isn’t. In my address book are 1020 people. If each of them give $20 we can fund this. Not a lot to ask.

If you’re interested – follow the link and donate on the Unicef site. A tax receipt will be issued via email by Unicef for each donation received.


Strategy Web Analytics

My project approach

In the big picture of things, the web is not that old. But, like my son, it has grown leaps and bounds in a short time and is maturing at an exponential pace.

When I first started building websites back in the 90’s and met with a prospective client I’d take them through four questions which I would throw in a simple 4-column chart that looked like this:


The questions I would ask were:

  • Who’s your audience?
  • What are your objectives for each of those audiences?
  • What strategies are you going to use to fulfill each of those objectives for each of those audiences?
  • What tools/content/functionality do we need to build to address each of those strategies?

I’d take them through each question sequentially, progressively building on the prior answers, and captured everything in the clients words. In the end would have a pretty good idea of what I needed to create and could go away and build out a quick estimate and proposal knowing I had everything I needed. This approach worked pretty well (and still does) and made even small clients stop and think a bit about what it is they wanted to create and spend their money on.

Over the years I’ve adapted that model and here’s how it looks now:

Define the audience

Step back for a minute and consider who you are you building this for. Who are they? Can you describe them, tell us about them? What are their needs? Why this audience instead of another? Make some basic notes, flush out some user profiles or dive deeper and build out living, breathing personas. The better you understand “who” the better off you’ll be.

Define/discuss the business outcomes

Wikipedia defines a business outcome as “an observable result or change in business performance possibly supported by transaction-based metrics, resulting from an event or action“. IMHO every project should have a clearly articulated outcome, but most don’t. And, most companies will have a hard time explaining “why” they are spending $X with you to do X work. Outcomes bring honesty to the table and accountabilities for all involved. It’s scary for many, but worth it.

Business outcomes can be quantitative and defined in terms of their business value (e.g. create new revenue or value, save money or time, lower risk) and expressed explicitly such as “increase our sales revenue by 10%”

They can also be qualitative (e.g. general satisfaction, staff more happy and engaged).

Capture these in your client’s words. If you have some baselines you can capture (where you are now vs. where you want to be) all the better. If you can do them for each of the audiences, better still.

Plan your strategy

So what specifically are you going to do in your project to meet the business outcomes? Involve your client in this process (don’t do it in a silo). Talk about content, functionality, communication, marketing, governance, change management. Go deep. Do it as a team and get everyone involved.

Define the metrics

Your metrics should be measures that help you understand how you are doing against your outcomes. Ideally these are KPIs (rates, ratios, averages, percentages). I tend to avoid raw number because they don’t provide context.

Determine how you will measure your outcomes

Often organizations miss this part and define metrics that they have no way to measure. It’s a good check/balance for the metrics you have defined. Try questions like “what data will be used to validate the above metrics?” Where does the information reside?

Spread the word

Make sure everyone knows the outcomes and strategy. Make big signs, put them in every presentation, live and breath your client’s desired outcomes and hold each other accountable for addressing them. When you veer off-track, reconnect and review what guided you from the start.

Check to see how you are doing

Check-in frequently to see where you are at in relation to your goals. Re-visit continually during the project. Do a post-mortem at the end of your project and check again. After you’ve delivered, check back and re-evaluate. Look for opportunities to go back to the above business outcomes and deliver more value. This is where I loved being a web analyst and digging into the data, identifying opportunities to improve and make things better.

Talk about your successes and failures

Be up-front about where you’ve delivered on business outcomes and shown value. But also don’t shy away from understanding where you’ve missed the mark (as you will miss sometimes) and leverage that knowledge to do better in the future.

Hope you found this helpful and interesting. I’d love to hear your thoughts on this and your approach.

Web Analytics

15% off Predictive Analytics World

Readers of the “Benry Blog” can get 15% off Predictive Analytics World. Just mention the following discount code when registering: benrypaw09

Web Analytics

EMetrics Canada

The 2009 EMetrics Canada conference has been announced and runs March 29 – April 1 in Toronto. This will be the second year the conference has been in Canada and I’m looking forward to seeing the lineup Andrea Hadley puts together.

For more information visit:

Web Analytics

MS Word as a Blog Posting Tool

I’ve been playing with MS Word (Windows) as a blog posting tool. It’s pretty easy to get it up and working too.

  • Create your blog post in a Word document
  • Go to Office Button and select PUBLISH > BLOG. Word supports Windows Live Spaces, Blogger, SharePoint, Community Server, TypePad, WordPress and others
  • Then publish your post

One thing nice about this is the ability to create posts offline (like when you’re at a conference or without internet access).

Web Analytics

Many Eyes

If you haven’t toyed with it yet, Many Eyes is an interesting tool for sharing and visualizing data buried in spreadsheets, tables or text. Created by scientists at IBM, it’s a compelling for visualizing information, discovering unexpected patterns or insights. I haven’t had the chance to run some web analytics data through it, but I suspect it would be fun to do. It’s a nice compliment to Wordle another text based tool, but more flexible.

Analytics Vendors Web Analytics

ClickTale upgrades

The folks at ClickTale have upgraded their app to include some new features. Included in the changes:

  • Form analytics (read about form analytics on their blog post)
  • Improved recording and playback so you can now see all visitor actions inside online forms including mouse movements, keystrokes and interactions with controls such as drop-down lists, check boxes, radio buttons, etc.
  • New aggregate cross-site reports: “Mouse Moves vs. Mouse Clicks”, “JavaScript Errors”, “Tags Report”, “Page Open Period and ActiveTime”, and “Page Load Times”
Looks like a nice upgrade.
Web Analytics Web Analytics Association WAA

WAA: web analytics championship

Start your engines!

The WAA has launched it’s Web Analytics Championship. It’s an opportunity for Web Analysts to have some “analytical fun”, share their knowledge, and get a big prize! Dive into the data via the Google Analytics account and see what insights you can come up with.

First prize is an eMetrics Marketing Optimization Summit pass including hotel expenses (nice) and submissions will be judged by Avinash Kaushik, Neil Mason, June Dershewitz (three of the nicest people in the WA space).

SEO/SEM Web Analytics

LunaMetrics: Google Analytics Training Day

LunaMetrics is holding a Google Analytics Training Day in New York City on June 4. Robbin Steif and her team, IMHO, are one of the best Google Authorized Consultants around (of course I’m biased since I’ve used their services in the past — but they really are the best). If you haven’t booked already, head on over and register.

Robbin kindly agreed to a little mini interview today about the event. Here’s what she had to say:

Scott: $285 — that’s a cheap 1 day event. Are spots filling up quickly? Will be nice at the Harvard Club — guess that would be because you’re a Harvard Alum?

Robbin: Spots are filling up, although I asked for a really large room, so I think we are ok, at least for now. As far as the Harvard Club goes — it’s not directly because I am an alum, it’s because — as an alumna — I have been there many times, and I know what it’s like. (In between college and business school, I worked for Harvard Magazine, and took prospective advertisers there for lunch.) I didn’t have to go check it out, which was a real bonus.

Scott: Would you describe this training as applicable for all GA users? Beginner,
Intermediate, Advanced (on your site you note that if you haven’t worked with
either GA or other web analytics packages that this might not be the right event
for you).

Robbin: Yes, it is for all levels. We are going to work very hard to make sure that all levels are happy because the person who is just starting has different questions from the person who has been working with it for a couple of years. As far as the person who doesn’t have GA – I think s/he needs a marketing seminar (“Why GA is good for your organization,”) and this is straight education.

Scott: Why did you make the decision to break off into tracks in the afternoon rather than running a second day around these more advanced themes/topics?

Robbin: That’s so interesting, that you think of these as more advanced. In fact, one track is marketing (“How do I use all these data?”) and the other is implementation (“So how do I enable my users to take their cookies with them from my domain to my third party shopping cart?”) Those are often two different people who care about those topics.

Scott: If you had to think of 5 major issues that most people implementing GA run into what would they be? How will your training day resolve these common problems?

Robbin: Well, there are issues around using GA, and then there are issues around implementation. The big marketing (user) issues are “What does all this stuff mean?” and “How can I leverage these data to make my website better?” The #1 implementation issue that new users run into (after they get past, “Why don’t my data show up immediately?” and “Why can’t I see last year’s data?”) is, “How can I make my email marketing and other campaigns talk to my Google Analytics?” As you can see, the marketing (using) issues are very global and continue from the beginner on to the expert, and the implementation issues are very specific and get solved quickly. They are two different beasts.

Scott: Who from LunaMetrics will be on-hand for the training day? (you have such a talented team there)

Robbin: This is the time for false modesty, right? But seriously, Traci Scharf and John Henson will be there, and I will too. We also invited Megan Kiel from the Google Website Optimizer team, and she is going to speak on that topic in the afternoon.

Scott: Any plans to make this training available again in the future? (I can’t make it, others may not be able to either)

Absolutely. But Scott — we’ll do it only if you can make it. Next time, I will clear it with your calendar first…

Scott: Finally, will Avinash be making an appearance or will you be able to bask in your own glow as GA gurus? (ha, ha, ha)

Robbin: Well, he knows how much I love him and wish that he would be joining us, but he is going to be somewhere in Europe at that time.