See you at Emetrics

By Jason Carmel | May 2, 2008 2:14:29 PM

ZAAZ will be hitting emetrics hard in San Francisco. Here's where you can see us:

On Monday, you have two choices at 11:15:

Monetizing Site Behaviors with Jason Burby and Nigel Morgan - If you aren't familiar with what monetization is or you are familiar with it but haven't quite figured out how to make it happen for you, then check this session out. This is also a good place to stalk Jason B. if you want him to sign his book for you. (I have mine on our coffee table).

Creating a Web Optimization Culture with Judith Pascual - Judith will explain in her own inimitable way how to start a culture in your company that relies, nay, depends on analytics and optimization. This is a good one to catch if you're the metrics champion at your company and have run into friction getting your colleagues (or bosses) to see the light. And it's essential to catch if you are a one-man analytics band (gender neutrally speaking, of course). Judith will show you how to get the love you so richly deserve.

And then there's Lunch - That's right, ZAAZ is buying lunch on Monday. If that wasn't enough, you'll also be able to hang with our CEO Shane Atchison. I saw him open for Death Cab, and he was awesome.

On Wednesday, you have another couple of opportunities to catch ZAAZlings in action:

At 11:15, Jason Burby's hitting it again on a panel discussing Social Media Metrics. This is pretty advanced analytics mojo for how you can track the value of an interaction (e.g., a blog thread, a buzz campaign) as opposed to a conversion. Heavy stuff for the experts out there.

And then at 4:00, yours truly, winner of the 2007 award for "Most Likely to be Called by People who Really Wanted to Talk to Jason Burby," will be presenting Starting from Scratch - Effectively Using Kittens for Optimization and Usability. This is a fun topic that seeks a happy place where Usability experts and conversion-driven Optimization practitioners can make websites far better by working together. Should be huge.

We'll also be attending tons of sessions (we are all, after all, students of the craft) and are generally down for a drink or seven at the hotel bar, so please introduce yourselves if you bump into us. If you want to find me to buy me that beer I've had my eye on, or to talk shop or throw a challenge our way, don't hesitate to email me at jasonc at zaaz dot com.

See you next week.

Asteroid to hit an Earth near you soon!

By Nigel Morgan | Apr 17, 2008 8:38:26 AM

We had a fabulous story emerge this week about a 13 year old German boy “outfoxing” NASA. Apparently NASA has claimed that the asteroid Apophis has a 1 in 45,000 chance of hitting the earth on April 13 2036. Now this budding young German scientist, Nico Marquardt, claimed that the real probability is closer to 1 in 450 (apparently he added-in the possibility of the asteroid hitting an Earth orbiting object and being redirected towards the Earth). Clever lad that Nico; I am sure his inquisitiveness will make him a fine web analyst or even optimization manager at ZAAZ one day, if, of course, astrophysics turns out to be a bit of a fad. Nico if you’re reading this, get in touch, we can probably slip you some allowance up front!!! Anyhow, luckily for all of us, it turns out NASA was right after all and on that night in April many moons from now, we should all be able to sleep quite safely. Phew!

This amusing story of Nico shows how important it is to get it right when dealing with numbers, and how a small error or incorrect assumption can magnify its way through an analysis to have eye-opening consequences (by the way Nico did not consider the fact that the polar orbits of the near earth satellites would take them away from the area that the asteroid would pass through at the given hour). Still, bloody good effort there Nico, we at ZAAZ applaud you.

While most data I have ever had the pleasure of analyzing hasn’t had the same impact (excuse the pun) as the rather sobering “asteroid hits earth, everyone dead” gravity (there I go again), it has been important in its context of marketing superiority over one’s adversaries.

The ubiquity of data and the ability to statistically prove that one design, call to action, promotion is more relevant to your customers than another brings an exciting, unprecedented era in direct, one-on-one marketing. However it also brings its own set of challenges that any manager must appreciate in order to be successful. Focusing on the right data sets, being accurate in your measurement and building optimization projects with statistical precision are all areas that require a level of expertise. Knowing the importance of data driven decision making is only the first step and having good intentions when launching a data driven program does not mean that success is assured (I think the tool vendors have to take some of the blame for this by pushing the “so easy a cat could do it” sales messaging…this is simply not true). The reality is that website marketing managers often do not have the off-the-shelf expertise to manage a successful analytic or optimization program.

So whenever undertaking any analytic or optimization project, be absolutely sure you understand what numbers you need to look at and how to build a statistical or analytical plan for your project success. If this all sounds daunting and outside of your comfort zone; ask a professional for help. While there are not many of us around, we are relatively easy to find. 

However, if you decide to go ahead with your project and do not have the basic analytics and statistics covered, you may have explain to your boss that the project money has gone and you have nothing to show because you designed the experiment incorrectly, made mistakes setting up the tool, or did not have a clear statistical or analytic plan for the project or program.

Of course that is a worse case scenario but unfortunately we see it with startling frequency. Now, if that happens, proverbially speaking, it will probably feel that Nico was right all along and the Apophis asteroid has hit the Earth, or your part of it, after all.

Behavioral Targeting and the FTC

By Jason Carmel | Apr 8, 2008 2:01:04 PM

The Federal Trade Commission has a call out for comments on their proposed Principles for Behavioral Advertising. There are a few comments posted already- one by Google (well, duh), a few by concerned citizens, and the obligatory post or two by concerned citizens who can't conjugate verbs. Ah, democracy.

I just submitted comments on behalf of ZAAZ and am trying to unpack the experience.

When I read the proposed principles, I had a fleeting understanding of what it must be like to work in our User Experience Group here at ZAAZ. Our crack UX experts (UXperts?) often field requests from well-meaning potential clients to "help us be more web 2.0" or to "build a community online for users." Instead of jumping in and architecting something, the ZAAZ team always takes a step back, and reframes the issues to make sure that the proposed idea really solves a problem that the customers have.

I get the feeling that the FTC could greatly benefit from this approach. To its credit, the organization has been paying attention to the emergence of behavioral targeting applications and the blowback by folks who are creeped out about their every Internet move being tracked in order to present (sometimes eerily) relevant ads. Maybe it's time now to take a step back and understand what problem we are trying to solve and whether there are already mechanisms in place to solve them.

My overall thoughts:

Spend More Time With Definitions - At the end of the day, I'm not sure there is a clear enough distinction in the Principles between ad-side behavioral targeting and site-side behavioral targeting. It feels like the FTC is talking exclusively about the former, but they need to make that patently clear before they get all up in our grill (as it were). Should I see my name and address on an ad for mortgages on a site I've never visited before? Probably not. Should a site with which I have an existing relationship, with whose Terms and Conditions I've already agreed, be able to offer me a discount on a product they just released that relates directly to one I've already purchased from them? Almost definitely.  Both are behavioral targeting. That needs to be sorted out before anything.

Let the Market Decide What is Acceptable - It does not take a great deal of effort to troll the annals of Internet history in search of a business that tried to get a little too forward with its customers' privacy and got a black eye as a result. The Facebook Beacon fiasco is one example that jumps to mind where a company was neither very open nor very clear with how it planned to treat its customers. The resulting poop-storm from the media, industry pundits, advocacy groups and customers regarding a "feature" that could have probably used a few more minutes in the oven engendered an almost immediate about-face from the company, and taught anyone who witnessed it an object lesson in how customers own the relationship. I think the FTC should view this an ideal model. I don't mean to get all "invisible hand" on you, but the best way to keep a company in line with regards to privacy is if they know that they could lose a bunch of customers by misbehaving.

Use Existing Laws Where Possible - I'm all for law. Seriously. "Yaaaaaay Law," as far as I'm concerned. There are gobs of existing laws that pertain to privacy with regard to specific types of information that can and can't be shared (financial, medical, child-related, etc). We should probably us those as stepping stones before enacting new (and, dare I even suggest, potentially conflicting) rules. Otherwise, things will get messy fast. Plus, basic contract law instructs us that a company that changes its privacy policy without informing customers has basically breached its contract and should be liable for any damages that occur as a result.

No Double Standards for Online vs. Offline - Let's be realistic here. You're grocery store tracks more data about you than any practitioner of online Behavioral Targeting could ever dream. Same with your phone company. And your airline of choice. And your credit card. If the practice of targeting ads should be regulated, it doesn't make sense that we are only considering online participants.

If the FTC rephrased the request for comments using some of the above recommendations, I'm confident that both businesses and consumers would end up in a happier place. For my money, I think the FTC could do the most good right now by helping consumers understand the opt-out options that already exist for Behavioral Targeting, such as the one made available by the NAI.

Please feel free to review the full text that ZAAZ submitted, conveniently made available on the FTC website (note: be forewarned: the document is nine full pages of formality, with footnotes and everything. Read it with someone you love.). And by all means, don't be shy about giving the good people at the FTC your own opinions. Tell them that Behavioral Targeting is our only hope for civilization as we know it. Tell them you think Behavioral Targeting is a sign of the impending apocalypse and should be punished by extreme pain. Or tell them that you agree wholeheartedly with the witty, urbane, and devilishly handsome distinguished gentleman from ZAAZ.

You have until the end of the week.

Optimizing for Micro Conversions

By Jason Carmel | Mar 26, 2008 2:40:06 PM

Avinash has a great post about measuring micro conversions that lead up to a major conversion (usually a purchase). He summarizes thusly:

But people don’t just come to your site to Buy. They are there to Research products and services (and buy offline). They are looking to get Support. They are there for looking for Jobs. They might be there to look at your latest Blog Post, etc.

Make each one of those your Micro Conversions....and then [use] your Web Analytics tools to measure success.

I have two comments:

Comment the First

I don't know when this happened, but Avinash has morphed into one of those people who can be identified exclusively by one name, without any risk of confusion. I love that. He is the Bono of Web Analytics.

Comment the Second

Let's think about what micro conversions do- they provide success metrics that sit in front of a final purchase for a site. Properly tracked, they can illuminate the key triggers that lead to a purchase, and if you're really smart (and, let's face it, you are), they can back you into a monetization model that will actually put a dollar value on each micro conversion event (which is something that ZAAZ sees as essential to an online presence). Monetizing these micro conversions provides a useful way to roll up the entire value of the web channel efforts, but even more importantly, the values you ascribe will help prioritize customer activities and the resources that support them. Otherwise, how do you know that a blog is driving more value than a video product demo? Or whether the user review section is more critical to a purchase than a web specials page (as two of, say, a gazillion examples)? And if you can't answer that type of question, how do you know which effort deserves the budget?

Of course, once you jump on the monetized micro conversion bandwagon, the testing possibilities explode out. Testing micro conversions that sit before a final purchase has the additional benefit of (usually) seeing more traffic than the deep funnel (i.e., purchase) macro conversions do. As such, when you engage in testing micro conversions, you can increase the speed at which you run tests and the number of variants that you include within each.

When you report on results for micro conversions, don't forget to multiply the additional conversions by the monetized value so you can present a real dollar value that your test has added. That is, instead of saying this:

"Our test resulted in 12% more visits to the "We Heart Pickles" video."

Say this:

"Our test resulted in 12% more visits to the "We Heart Pickles" video. Each visit to the video brings in about $0.60 of value through eventual purchases of pickles (Jars of pickles cost 6 bucks, and 1 out of every 10 visitors to the We Heart Pickles video purchases a jar). The 12% increase means that 10,000 more people per month saw the video, which resulted in an increase in monetized value to the site of $6K/month, or $72K/year, which averages out to about 12,000 more jars of pickles sold every year."

That's a lot of pickles. And a far more compelling story to tell.

As an intellectual exercise, can you think of any site that doesn't have a micro conversion? Can you think of what the top 5 most important micro conversions for your site are?

What would you say to Stephen Hawking?

By Nigel Morgan | Mar 10, 2008 5:27:29 PM

We work in an environment where intellectual horsepower, a curiosity for the unexplained and a fluency in numbers exist in melodious hegemony. As professionals in the field of data decision making we ask questions, form hypotheses and execute with statistical precision.

I have found throughout my career that people who work in this field willingly express their ideas, opinions and “expertise” with other like minded souls whenever presented with the chance. I am guilty as charged with these offences, perhaps more so than others.

Just yesterday I had one such opportunity to commence what would have been a profound conversation quite literally about life, the universe and everything. However, as I tried to create a funny, insightful, intellectually pertinent opening comment, I found myself disturbingly unable to find anything relevant to say. Worse, within a minute the moment had passed, I had not said a word to my quarry and I had missed the opportunity altogether.

So what was this event? Well, following a particularly moving rendition of Puccini’s Tosca at the Seattle Opera, I had the pleasure of coming face to face with the esteemed Professor Stephen Hawking. While any individual with the aforementioned skills in analytical prowess and intellectual curiosity would leap at a chance to be in that position, I found myself an unintelligible blob.

This chance encounter and the subsequent cerebral impasse have befuddled me for the past day. From the moment my mind drew a blank I have lamented the missed opportunity to engage with the greatest mind of our time. Instead of being too cutesy with my message, maybe I should just have asked him if he enjoyed the performance (we did just share an experience for the prior two and a half hours). Maybe that would have enabled me to engage in a relevant, initial conversation with the revered professor and thereafter move on to a more intellectual and significant topic?

Finally, now that I have determined that I should have differed in my approach, I have come to a premise that should be applicable to you or any online marketer. Being quick, direct and relevant to the person you wish to engage (even if not supplying a core message) could be a better way to connect than brainstorming for hours in a creative loop, delivering something potentially irrelevant and missing the opportunity altogether.

So, before you undertake your next great marketing endeavor, think about who you are trying to reach, and how you can engage on a topic relevant to them rather than you. Why not set aside a couple of minutes and take this rhetorical challenge: If you had two minutes only in an elevator, what would you say to Stephen Hawking?

Omniture Summit 2008

By Jason Carmel | Mar 7, 2008 7:04:03 PM

Say what you will about Omniture, but they sure can throw down a conference. We had around a dozen ZAAZlings at the 2008 Summit, where everybody in the Omniture customer and partner world comes to see and be seen. Imagine Sundance, but with visitor segmentation and monetization models instead of Robert Redford.

Roughly 2000 people poured into Salt Lake City for the event in search of technology, news, partners, customers, jobs, and vendors, but mostly for ideas. Everyone I talked to at the conference had a problem, slightly different from everyone else's, that required some cross-pollination of ideas to resolve. The keynotes, etc., while quite good in and of themselves, were second fiddle to the value of simply getting that many people who deal with online analytics and optimization into one square block for the better part of a week. The team at Omniture should get a medal for setting that up and for running it with grace and style.

And of course, the Flight of the Conchords concert kicked seven different kinds of ass.

The big reveal in the Optimization world was the rebranding of Offermatica and Touch Clarity into Omniture Test&Target. I wholeheartedly agree that both products are fantastic in their own ways, and that there is substantial potential for full integration of the testing and behavioral targeting components, but my world was not rocked by the announcement. The rebranding at this point is more an indication of an intent to integrate the tools- not an announcement that it's been done. It's an important step, but only a first one.

Jason Burby and I had a great time talking about Actionable Analytics to what must have been about 400 people who were patient enough to switch rooms (the first one only had about 80 seats) AND to wait while the tech folks reloaded the presentation and found microphones for us.  Hopefully, the concepts of monetization and optimization, as we use them at ZAAZ, were ideas that helped some people solve address the problems they brought with them to the conference.

I Want to Test You Like an Animal....

By Jason Carmel | Mar 3, 2008 5:05:46 PM

Nine Inch Nails front man Trent Reznor should be applauded for experimenting with his own work to revamp what appears increasingly to be a broken business model for musics sales. Following other artists (most notably, Radiohead), Reznor is offering his new 36-track album Ghosts I-IV in five formats at five different price points, from free (first nine tracks downloadable) to a $300 signed, limited-edition package that includes multiple CDs and vinyl (if you're into that).

This decision must have been very difficult for Mr. Reznor. How did he know which five offerings would resonate most with his listeners? How did he know he got the most value out of each sale? Absent any data going into this, it must have felt like a leap of faith.

There are some decisions that legitimately require a leap of faith. For example, when you go to a hair salon, my assumption is that you pretty much have to sit in the chair and pray that when you said "something fun with reddish highlights," to your hairdresser, he or she hasn't recently eaten at McDonald's.

Fortunately, most business decisions don't require that much of a leap if you're smart about it. You have gobs of modeling and competitive research at your disposal, plus focus groups and beta testing if you really know what you're doing. Even the aforementioned Micky D's would occasionally test market some of their more creative culinary offerings, just to see what the reaction would be before fully committing worldwide to something like the McHaggis.

I assume Mr. Reznor is a smart cat, so I wonder what sort of testing he did prior to packaging up his new tunes. Did he send an email to his fan base with different offerings and price points to gauge optimal response of pre-orders? If not, is he looking at the results as they come in (he's running Google Analytics, BTW)? And if he doesn't like what he sees, will he test different options on his site, or just dismiss the idea as one that never had any legs to begin with? As both an optimization practitioner and a music fan, I hope it's the former.

Mr. Reznor, if you're reading this, shoot me an email and I'd be happy to walk you through a few ideas I have in exchange for some tickets the next time you swing through Seattle. Your music is well represented on the shared iTunes at ZAAZ.

Speaking of road shows, ZAAZ will be on tour for the rest of the week at the 2008 Omniture Summit in scenic Utah. Stop by the booth if you're there. I'll be presenting with rock-star Jason Burby on Thursday, so check us out.