24 posts categorized "tracking"

June 30, 2014

No Diving Allowed: Putting Performance First to Make Sure Your Projects Start Right

Starting a new client-agency relationship can be very exciting. It isn’t uncommon for a new pairing to be jointly celebrated with announcements given to a cheering crowd, toasts, and cake. Once a company has decided to invest in a new digital marketing project, those responsible for its completion are under considerable pressure to speed up the development cycle and start getting results. The primary result project managers seek is the conclusion of the project. However, in all of the excitement about the new project and the rush to get it done, design teams must not neglect the first step that will make it possible to determine the project’s success in the first place – the business objectives against which it will be measured. Going forward with design without a structured measurement model is like diving into a pool without knowing how deep the water is – and not being terribly confident that you even know how to swim.

The beginning of a digital project is critical. Take, for example, a website redesign project. The motivation that initiated the request often is no more complicated than an executive team’s general dissatisfaction with the look and feel of the company website. They may have a sense that the site is not “user-friendly,” and they might even have a few anecdotes to support this theory.

This seeming lack of accountability to concrete, measurable results may seem to be a blessing for the team responsible for building it, but it produces a lack of clarity that is actually quite detrimental – and costly.

How costly? 

According to an article written for the Institute of Electrical and Electronic Engineers (IEEE), avoidable errors in software design waste billions per year. Some of the more egregious examples include a $400 million purchasing system abandoned by Ford (2004), $3.45 billion tax-credit overpayment by the UK Inland Revenue (2004-2005), and the explosion of a $350 million rocket by Arianespace (1996).

In fact, claims IEEE, the total estimated yearly cost for faulty software is enough to

“…launch the space shuttle 100 times, build and deploy the entire 24-sattellite Global Positioning System, and develop the Boeing 777 from scratch – and still have a few billion left over.”

The scope of the losses is difficult to put into context when accustomed to considerably smaller budgets for digital projects, but since digital marketing consumes 17% of a typical marketing budget for the year, it had better not be wasted. Regardless of how meticulously it was conceived, no effort is without risk. However, the IEEE study suggests twelve ways to mitigate that risk.

Two of them deserve special attention:

  1. Set and document realistic and meaningful project goals, and
  2. Craft well designed project requirements.

Think about that for a moment. Two of the major reasons software projects fail (and a website is a form of software) – costing billions to the organizations who initiate them (passed on to everyone else through higher prices, taxes, or debt) – have to do with the act of defining the thing the team is building. Analytics guru Avinash Kaushik (Web Analytics an Hour a Day, Web Analytics 2.0) put it this way

“The root cause of failure in most digital marketing campaigns is not the lack of creativity in the banner ad or TV spot or the sexiness of the website. It is not even (often) the people involved. It is quite simply the lack of structured thinking about what the real purpose of the campaign is and a lack of an objective set of measures with which to identify success or failure.”

 The problem, more often than not, isn’t the lack of a definition, but an abundance of them. To make sure your team working toward the same end, your company needs a well-designed, structured measurement model that follows the project through each stage of the development cycle. This will keep the entire team focused on the objectives and guide the user experience strategy.

 There are 5 steps to building an effective measurement model.

  1. Define Your Objectives – There are three outcomes that can be defined here. Only three. Increase revenue, decrease costs, and improve customer/audience loyalty. Everything else is in service of one of these three. These objectives must be specific, measurable, achievable, and they must deliver value to the organization.
  2. Assign Goals for Each Objective – Goals are the ways to achieve the business objectives. They represent the strategy (not the tactics) the campaign is delivering on. To demonstrate this, consider that one of your goals may be to generate leads. A company can debate the way to get that lead based on effectiveness and resources. Is a simple contact form sufficient, or should they require contact information to download a whitepaper?
  3. Identify Key Performance Indicators (KPI) – A KPI is just a metric on steroids -- one that ties performance directly to your business goals. There are a number of metrics that it’s a good idea to track, but if a metric you feel is critical to your success doesn’t meet that criterion, either the metric is wrong or the model is.
  4. Set Targets – These are specific, numeric goals for your KPIs. In order to know whether the target is achievable, you will need to know or be able to find out what it is worth to your organization. Get the finance department involved. Look at your historical metrics. Use one of the lifetime customer value calculators floating out there.
  5. Segment – This is the key to getting actionable intelligence out of the data. It allows you to see who your most valuable consumers are. It allows you to see which advertising is most effective, under what circumstances, and how long it takes to convert and what the most common paths are.

It seems paradoxical, but the best way to improve your business quickly is to slow down. Do not rush right past the definition stage into the project design faintly hoping that it will provide executives comfort that the project is moving forward. Instead pace the progress appropriately to help them understand the activities that are necessary to improve their business. This will save you time and money in the design process by eliminating rework as each team member attempts to identify all of the incongruous assumptions on the fly, and will increase your opportunity for success for the entire project. 

January 13, 2010

When can a comma cost you $2 million?

Little details matter. Ask Rogers Communications, Inc.

In 2006, this Canadian company witnessed firsthand how a single comma in a contract could cost them over $2 million.

What they thought they signed:
The agreement “shall continue in force for a period of five years from the date it is made, and thereafter for successive five year terms unless and until terminated by one year prior notice in writing by either party.”

What they actually signed
The agreement “shall continue in force for a period of five years from the date it is made, and thereafter for successive five year terms, unless and until terminated by one year prior notice in writing by either party.”

The second comma changed the meaning completely. Whereas Rogers Communications thought only the subsequent extensions could be terminated on one year's notice, the clause created by the comma meant that the initial 5-year agreement could be canceled by either party. Consequently, the rates they were obligated to pay shot up immensely within the 5 year period they thought they'd have the prices locked in. (Read the story)

Details can make or break your website
Hopefully you have good lawyers who will, among other things, indemnify you in case someone maliciously uses your software or website to build weapons of mass destruction. Like Apple's lawyers did with iTunes. (Read iWMD: Why No One Reads License Agreements)

But even with that important detail taken care of, the little details matter in user interfaces, as well. And failing to pay attention to them can be the difference between success or failure.

  • Should that call-to-action be a button or a link?
  • Should those calls-to-action be together or separate?
  • Should the calls-to-action be of equal weight, or should one be given greater priority?

How you answer those questions depend on what it is you're trying to accomplish and what people are expecting to find. But on a high-volume or high-stakes site, if minding the details can improve your conversion metrics by just 5-10%, it could be the difference between profitability and a money-leaking ego booster.

The Web is your petri dish
If at all possible, don't rely on experts to tell you that something has to be one way or the other. Test early and often. Don't be afraid to try new things.

Work diligently on the details. In bits.

  • Is the headline effective?
  • Is the language on the button inviting?
  • Does the button look imminently clickable?

Let the data speak for themselves. You may want experts to design the page and the test, but you don't need an expert to know that a 15% conversion rate is better than a 10% conversion rate.

There are plenty of cost-effective experiments you can run to help you get the most bang for your buck, including A/B split testing and informal low-cost usability tests.

However, the characteristic you must first have is a willingness to fail. Because only through failure can you foster a willingness to search for the problem and design experiments to help you improve. - Cam Beck

June 11, 2009

Keep it simple

Ist2_440585-information-overload The transition from traditional media to digital media is a lot like going from a desert to a rain forest. Although there are quite a few ways to measure traditional media, many times it's not cost effective to pay firms to get the data. It's all a bit suspect too because you're forced to use sample groups that may or may not be representative of the rest of the population.

Online media is infinitely measurable. Geography, activity, technology and in some cases even behavior before and after the click is measured in a way that floods an analyst with data. Those of us who have built our careers within the digital space love to pour over the data and talk about all types of information. The problem is that not all of it is relevant.

I love this recent article from Mark Walsh from the OMMA conference on Avoid 'Data Diarrhea'. Although I wasn't at the conference, I can relate to what must have been said:

"Data is useless unless it becomes currency for making a decision," he said. To make sure online measurement is on track, he advised advertisers and analytics professionals to be sure they are asking the right questions, have the right expertise and the right mindset"

It's important to begin with the end in mind. Understanding what metrics are important should be done at the beginning. Dash boards for digital marketing should resemble car dashboards, not the dash board for the Space Shuttle.

An important reminder.

- Paul Herring

February 27, 2009

Five in the Morning

A little while back, Steve Woodruff tapped me for continuing his excellent "5 in the Morning" series, which allows bloggers to highlight articles you might have missed. Below is my contribution.

  1. How to approach a problem by Inspire UX
  2. Is the beautifully simple Macintosh OSX interface the panacea of interface design? Not according to Bruce Tognazzini of NN/g.
  3. Anyone who has ever bought a bleeding-edge device and had difficulty trying to use it can surely appreciate this piece from the Onion via 90 percent of everything (language advisory).
  4. Business Intelligence is rated a #1 technology priority for CIOs, and #5 business priority (Perhaps this second number should be moved up).
  5. Steve Roesler advises us to look at presentations the same way Oglivy looked at advertising. Long live the big idea.

Thanks, Steve, for allowing me to participate!

Subscribe: ChaosScenario / Steve Woodruff's StickyFigure blog
Follow on Twitter: Cam Beck / Steve Woodruff

February 06, 2009

Blog, Facebook, Twitter and Myspace: Managing Your Profile and Hedging Your Bets

As they say, the only two things that are certain are death and taxes. The rest is open to interpretation. This includes the suitability of a candidate for any given position. Therefore, whether you're looking for a job or looking for a good person to fill a job, you're putting something valuable at risk -- either your time or your money. If job hunting is a gamble, then why not hedge your bets a bit?

The following is part of an exchange between banking guru J.P. Morgan and a member of a congressional committee in the early 20th century.

Congressman: Is not commercial credit based primarily on money or property?
J.P.M.: No, sir. The first thing is character.
Congressman: Before money or property?
J.P.M.: Before money or anything else. Money cannot buy it.

Long résumés and walls full of degrees and commendations have their place, but in the pantheon of qualifications, I wouldn't rank them the highest. Like Morgan, I would put a premium on character. In fact, here's how I would rank the order of importance.

  1. Character
  2. Intelligence (general and specific)
  3. Specific experience
  4. Education

Why you should be all over the 'net

The problem is that character and intelligence aren't easy to put on a résumé. Some people get by with listing experience (especially pro bono work they've done) and education, but that will only get you so far. Although they're often a good starting point, companies know the story doesn't end there.

And increasingly, they know how to use Google.

Knowing this, you really have only two choices:

  1. Ignore it and try to fly under Google's considerably effective radar, or
  2. Embrace it and influence it as much as you can

Number 1 may work only if you have no friends or a name like "Abraham Lincoln." In the first case, even if you try to fly under the radar, your friends may have other ideas about your desire for anonymity, and if you're not actively telling your story, someone else may be. In the second case... let's face it... Companies would be hard-pressed to find information about you when they have to navigate through all the information about that other guy.

Number 2, on the other hand, gives you an opportunity to tell your side of the story before anyone else does. If you're a private person and are uncomfortable with being "out there," you have the ability to moderate your level of personal disclosure.

If you're prolific, you can ensure prospective employers (or clients, if you're self-employed) can become convinced that you may have the qualities they seek in a candidate. If they don't value those qualities, they're probably not a good fit for you anyway.

That way, if there are any issues (fairly or unfairly) that call your character into question (Remember those parties you attended when you were 25?) they will be drowned out by the story you'd rather they see.

Just make sure your online persona is consistent with the way you want others to see you. - Cam Beck

December 02, 2008

Advertising to Calculus Students: The Standard of Proof

Try these math problems on for size:

  1. Tom Farber, a calculus teacher in San Diego, California, has a $316 budget for copies. He needs $500 to provide his students with the practice they need to master the course.

    True or False? $316 < $500
  2. Getting a larger budget from the school is not an option. There is no discretionary or R&D budget. What is the most sustainable way for Mr. Farber to solve this problem?

    A. Pay the $184 difference out of his own pocket.
    B. Keep printing tests as normal, hoping the $184 would be found somewhere between now and then.
    C. Reduce the number of tests. Use up only what his budget allows, possibly sacrificing the course's effectiveness.
    D. None of the above.

Assuming his methods really are effective, give Mr. Farber extra credit for choosing "D" and refusing to sacrifice the quality of his students' education. When faced with the prospect of a budget shortfall, Farber invented a way to make up for it.

Testadsx-largeHe sold advertising.

On tests.

And worksheets.

Even the final exam.

But is it ethical?

The managing director of Commercial Alert, an organization formed explicitly to prevent commercial interests from intruding on spaces they deem ought to be impenetrable, is worried that, since Mr. Farber raised nearly twice as much as his shortfall, this feat might be duplicated by other struggling schools.

Before jumping to conclusions, here are a few questions worth asking:

  1. Do Mr. Farber's extra tests make the course more effective at teaching Calculus?
  2. Can that effectiveness be duplicated by some other means, less expensively?
  3. Does selling any ad space, whatsoever, in any format, inherently corrupt the education process?
  4. Do ads inherently corrupt the students?
  5. Is it possible to corrupt students with advertising?
  6. What measures can be taken to prevent it?

When are sponsored ads on tests justified?

If the answer to #1 is yes and #s 2-5 are no, then by all means, sell the ads and let the kids learn.

If the answer to #1 is no, then it doesn't much matter what the other answers are, because the problem isn't with the number of tests, but with the teaching method.

If the answer to only #1 and #5 are yes, then it becomes necessary to answer #6 and to determine which option costs less: Forking over the dough for copies or investing the time, energy, and money to institute proper safeguards that prevent the corruption of the students or the process.

Who is responsible for finding out?

The teacher and the school have the duty to educate the children under their care. Thus, it is up to them to confirm the effectiveness and integrity of the system.

Once they've established a positive correlation between number of tests and better grades, they can determine if the better grades were caused by another factor, such as some sort of quid pro quo between the teacher and the advertiser.

Beyond that, the public can subject the process to what is called a "reasonable person" test. According to Wikipedia:

The reasonable person is a legal fiction of the common law representing an objective standard against which any individual's conduct can be measured. It is used to determine if a breach of the standard of care has occurred, provided a duty of care can be proven.

The standard performs a crucial role in determining negligence in both criminal law—that is, criminal negligence—and tort law. The standard also has a presence in contract law, though its use there is substantially different.

Therefore, if Commercial Alert or any other advocacy group believes the practice of exposing students to advertising is harmful in some way, they are obligated to not only assert, but provide a body of evidence that shows how what they claim could be true.

It is incumbent on them to do so, because they propose such advertising be eradicated by statute, and in doing so they eliminate a potentially effective resource for overcoming budget problems in a way that doesn't require compulsion.

Because unlike individual taxpayers, sponsors have the choice as to whether or not they will participate.

The standard of proof has to rise above the level of indignation - feigned or otherwise. Commercial Alert may have a point. But they may be blowhards. It's up to them to convince the public which category they fall under.

Likewise, Mr. Farber and his school shouldn't get a free pass, either. We shouldn't take on faith that his approach is necessarily better than one requiring fewer tests. But in this case, there should be a mountain of statistical studies already that suggest one thing or another.

All they need to do is cite them. - Cam Beck

March 13, 2008

PayPerPost: Friend or Foe?

Recently a representative for IZEA came to visit Click Here about opportunities afforded by PayPerPost, a service Paul has written about before. The goal we had in mind wasn't to explore opportunities for ChaosScenario (our relationship with our wide variety of clients would pose too many potential conflicts of interest to even consider), but to see if any of our clients could use the service. I've concluded very little, but I think I have an idea about the 3 main questions I need to consider. Before we get there, though, I have to provide a little background.

PayPerPost must be looked at separately for companies and bloggers. Here is how it works:

  1. Companies determine how many bloggers they want to reach, how much they will bid for the posts, what the requirements of the post will be, and who can be eligible to participate (readership, ranking, industries covered).
  2. Bloggers who participate in the program search for or are invited to write about something they want to write about based on company, industry, and/or the amount they are being offered. If they find something that strikes their interest, they sign up for it and write a post according to the requirements (word count, etc.).

Here are the rules:

  1. Bloggers are required to disclose that they are being paid for their contribution (this wasn't always so), and
  2. They must separate each paid post with at least one normal (unpaid) post. The subject matter doesn't need to be the same.

Based on how well the blogger did, and providing the blogger followed the rules, the paying company can rate the blogger's post, which affects the person's PayPerPost reputation. This impacts how many requests they get in the future.

Is it immoral? Not necessarily.
Empirically I see no intrinsic ethical problem with this arrangement. Each blogger must maintain the relationships with his own readers. Since bloggers must disclose that they are being paid for their work, the reader is free do decide if the endorsement has merit.

All people, in business and on blogs, who sacrifice their integrity will see their reputations punished by the marketplace.

But perception is 9/10 of the law
If a blogger's audience objects to their being paid for writing posts, then it is up to the blogger to figure out how to deal with that, or he will lose his audience. That is between the blogger and his readers.

However, justified or not, how other people (mis)behave affects others' perception of us.

So it's worth it, as a blogger and as someone who needs to make recommendations in the best interests of my clients, to analyze this development in the marketplace.

Beware of myopia
It's easy to sit here from our pedestals, as marketing bloggers and media critics, and think that everyone else who blogs thinks about blogs like we do.

The truth is that people consume media -- including blogs -- in all sorts of ways, for all sorts of reasons. I try to branch outside of my circle continuously to remind myself of this fact, and I've benefited (profited?) from the experience.

One blogger claims that she earned $14,000 in one year from PayPerPost, and that her traffic has actually increased since she started (she draws no conclusions about causality). That's not a king's ransom, but it does show how a blogger can be effective using the service, part time.

What's the big difference?
While considering this issue, I'm comparing these things.

  1. Advertising - If you accept advertisements on your blog, do you run the ads through a filter to determine a proper fit? Do you have to use the product or service you are advertising? Do you care that your display ads are largely ignored?
  2. Celebrity endorsements -  When Michael Jordan receives millions for pitching Nike, do we hold it against him that he got paid for it?
  3. Sports sponsorships - Stadiums are named after the highest bidder. Does this compromise the integrity of the coaches, players, or owners? Does it affect your enjoyment of the game?
  4. Show sponsorships - When Cali Lewis tells us that the latest podcast was "brought to" us by the new, lower-price Nokia N810, is she to be commended or condemned?
  5. Work - Most of us have jobs that pay us a salary and/or commission for providing advice to our clients. Is the quality of our advice diminished because we're getting paid for it?

What do you think about it? I'm really looking for the answer to 3 main questions.

  1. Is it ethical?
  2. Will it help or harm the medium, over time?
  3. With an average campaign cost of $19K that delivers higher click-through rates than traditional display advertising, is it worth it?

- Cam Beck

March 11, 2008

A College Degree Increases Gullibility by 36%

Huff_cover My college statistics professor repeatedly told me that no one would want to read an article by a statistician, but he was 95% sure that 97% of news reporters (±3%) always report statistics incorrectly.

Case in point: Periodically the media report on a study that "proves" earning a college degree will increase your earning potential, and there is plenty of data showing that people with college degrees actually do make more money. However, what they usually fail to tell you is that the study (the one calculated by actual statisticians) cannot conclude that the college degree caused the increase in income.

... This is a post hoc fallacy at its best. It says that these figures show that if you (your son, your daughter) attend college you will probably earn more money than if you decide to spend the next four years in some other manner ... Actually we don't know but these are the people who would have made more money even if they had not gone to college. There are a couple things that indicate rather strongly that this is so. Colleges get a disproportionate number of two groups of kids: the bright and the rich. The bright might show good earning power without college knowledge. As for the rich ones ... well, money breeds money in several obvious ways. Few sons of rich men are found in low-income brackets whether they go to college or not. - Darrell Huff, How to Lie with Statistics [emphasis mine]

A report on a new study released by the Harvard School of Medicine now says that earning a college degree will increase life expectancy. To prove this, they cite data that show people with a college degree lived 7 years longer than those with a high school education or less. 

Take note that 7 years is enough time to finish a bachelor's degree and still have 3 years to spare.

Furthermore, at least one article says those who performed the study ruled out income disparities as a contributing factor to differences in longevity.

Assuming the model they used to perform the regression analysis was correct, this doesn't rule out the possibility that degree-earners live longer because:

  1. They happen to be bright (and thus know that a healthier lifestyle increases their quality of life),
  2. They have access to medical care (which is not necessarily tied to income), or
  3. They are driven to excel (which may motivate them to pursue a degree in the first place).

In any event, to actually suggest a piece of paper is a magic pill that can improve health is so improbable that it is almost laughable.

By all means, get an education -- whether or not you also pursue a degree.

A good education can help you in any number of ways. But don't rely on misleading data to motivate you to do it, and don't presume that reporters are correct when they infer that your education starts or ends with -- or depends on -- a college degree. - Cam Beck

January 04, 2008

The Best Career Decision You'll Ever Make

To excel at any profession, in which but a few arrive at mediocrity, is the most decisive mark of what is called genius or superior talents. - Adam Smith

Professional development doesn't happen by accident. In fact, research says it take years (5-10) and thousands of hours (10,000+) to become an expert at something, and we can never stop learning.

That is true in marketing as much as any other industry, if not more so.

I know all the excuses for taking a break from learning, because I've used them:

  1. I've got kids.
  2. I've worked hard enough this week.
  3. My favorite program is on tonight.
  4. I just don't feel like it.

But unless you make a concerted effort to learn and grow, you will find yourself standing in a puddle of mediocrity while your peers and your competitors surpass you in knowledge and talent. This means not only constantly practicing in your field, but also having the foresight to acknowledge when we made a mistake and the courage to learn from it.

If you want to really give your personal and professional development a shot in the arm, find someone to collaborate with who will force you challenge your assumptions. Find someone who can help motivate you to experiment. This can be one person or a group of people; quantity isn't nearly as important as quality.

Unchain Your Brain
My brother Gannon has put together a very insightful series on lifelong learning and becoming an expert; No MBA is required. It is a long but worthwhile read.
Part 1: Learn to Learn
Part 2: Reading
Part 3: Collaboration

Refresh Dallas
If you're in the Dallas area and you have some influence over the design or implementation of any website, Refresh Dallas has a presentation that you must attend.

You needn't be a designer, a usability specialist, an information architect, or an experience planner. You only need be concerned about delivering results for your client.

Day: Thursday, January 10, 2008
Time: 7:00 p.m - 9:00 p.m.
Subject: Multivariate testing

"Learn how to test designs with almost infinite variations, using real users, for a dollar a day. Avoid arguments with designers by letting the public tell you which words and images work the best. Increase revenue and/or profits, and then walk into the meeting with the data to back it up. Bryan J Busch will walk you through the process of multivariate testing."

Click HERE to RSVP.

About Refresh Dallas
Refresh is a community of designers and developers working to refresh the creative, technical and professional aspects of new media endeavors in the Dallas area.

Hope to see you there! - Cam Beck

December 07, 2007

What iTunes Could Have Been

Red_logoThis is what iTunes should have been.

When someone directed me to a site promising to be the "social music revolution," I was pretty skeptical. I wasn't sure if I could use it, or if I would even have any inclination to. Anytime someone offers something they claim is "revolutionary," chances are it will be hard to use. What's more, I was worried that those who saw the value in the social aspect of the site would be inclined to ask for way more information that they needed.

To my delight and surprise, Last.fm is turning out to be a useful and simple to use music discovery tool. The sign-up process is one of the best I've seen, and the design is very attractive and mostly intuitive.

I downloaded the software and put it in my keyword search ("Christmas"), and now I'm listening to random Christmas music, commercial-free.

23668411959561531 And it isn't just the scrub songs, either. Right now Luther Vandross is bellowing the lyrics to "O Come All Ye Faithful," and I'm loving it.

Supposedly the system will take inventory of my listening behavior and offer recommendations for music I might also like (and even uses this data to help populate my playlist). As someone who has difficulty remembering the names of artists and finding songs that meet my eclectic tastes, this is a feature I really appreciate.

I cannot listen to every song, on-demand, and I shouldn't be able to without having to buy them... But I can listen to 30-second previews of many songs that are in inventory. That way I can confidently ascertain if the album is suitable for me.

The biggest problem I see is that there appears to be no option to use the interface to buy directly from iTunes, since that happens to be my favorite interface for buying music on-demand. Some songs are available for download from Amazon, but so far most of the "Buy" links take me to the page where I can buy the CD from Amazon - to have shipped to me.

If I want it immediately (which I would, since I'd already be in listening mode), I'm out of luck, or else I have to use a different interface (such as iTunes).

If the software allowed me to download the music directly and load it onto my iPod, this would truly revolutionize the way I buy music - and I'd probably buy a lot more of it than I do now.

Perhaps that's something still on the horizon. I sure hope so. - Cam Beck