Actionable Web Analytics

I've added study notes on Actionable Web Analytics, a book that means to be a primer on the topic for an individual who has been tasked with setting up an analytics operation in a corporate environment. It's possibly a worthwhile read for someone who's wholly inexperienced with the topic, as it merely grazes the surface of the topics in which a practitioner should be well-versed, but at the same time contains some very poor and misleading information.

The basic approach of the book is sound: that measurements and metrics pertaining to the behavior of Web site visitors should not be collected just because they can be collected - and that may of the commercial software packages perform an astounding number of utterly useless calculations, which produce intimidating reports full of statistics that are of absolutely no use to anyone.

Instead, some rationale is required to consider the wealth of data that is available, and identify the analyses that provide useful information - data that can be used to drive business decisions that will have a meaningful impact. And once collected and analyzed, the site operator should be willing to invest in taking action as a response.

But beyond this basic message, the book falls short of being a useful guide to the practitioner, in that it provides scant actionable detail aside of a few hypothetical examples. The reader is left to consider his organization, and the site it operates, to determine with little guidance what data he needs, how he might gather it, and how he might take action on the resulting analysis. It's not a good reference for anyone with even a few months' experience in the field - much of what is said is probably quite obvious.

Neither could it be a book that could be recommended to an executive for insight on how he might better leverage the analytics department within his organization. The basic level of detail might make it seem so, but the overt posturing and the attempt to claim for the Web analyst a high degree of control within an organization would likely cause an outsider to take a dim view of the profession, and mistrust analytics as merely a numbers-game that is rigged to serve the interests of the analysts.

Aside of that, there are a few particularly awful sections, where the author delves into ancillary topics in a superficial and often haphazard manner which is laughable to anyone who recognizes the inherent flaws, and dangerous to anyone who takes the information at face value. It's so much filler content that skims the surface of topics that merit much closer inspection from a more knowledgeable source.

And so, this book is to be approached, if at all, much in the same way it advises approaching the data that can be gathered from a Web site: with a suspicious and discerning eye, to pick through the meaningless and misleading data to find the few pieces of useful information that are to be had.

The Other 99.5%

The traditional practice of marketing fails to consider the persistence of human memory. Its goal is to onboard a very small percentage of a very large number of people. If two million people see an advertisement, and only 0.5% of them react to it, that's 10,000 sales (or leads, or whatnot) - and that would be hailed as a very successful campaign. But it fails to consider a very important question:

What about the other 99.5%?

From the traditional perspective, they just don't matter. At least you got your name out there; and maybe you can catch them next time, with a promotion that's more appealing to their specific interests, or catch them at a time when they're more likely to respond to an offer. Ultimately, the attitude toward non-respondents is "we don't care." And that's a terrible and costly mistake.

If your message was inappropriate, offensive, or misleading, they will be less likely to respond to your next message. And applying the same algorithm, if the message you send damages your brand to only 0.5% of the audience, that's 10,000 people who will be less likely to respond to an offer, even if it's better suited to their interests, and even if it catches them at a time when they're seeking to buy. Ever. You've damaged your brand and done long-term harm to your potential to gain new customers that may outweigh the value of the short-term gain of a sliver of a percentage of immediate sales.

I expect this is a bone of contention between brand marketing and sales promotion, though I haven't seen the notion get more than a passing mention. There have been suggestions that a company ought to follow up an promotional campaign with a brand audit to determine if they have done any collateral damage - but it always seems to be mentioned as an afterthought.

Collateral damage that, like so many dead civilians, is a minor embarrassment to those intent on winning a battle. The fact that victory was achieved is all that matters.

But the battle for market share is not entirely analogous to a military conflict. A promotional campaign is a tactical maneuver in a struggle than never ends in victory. The casualties you inflict are customers you will never gain, and who are likely to be more interested in your competitors' offer the next time they are in a position to purchase.

The "other" 99.5% should never be disregarded or dismissed.

The Wax and Wane of the Generic

On my last trip to Walmart, which is a blessedly infrequent occurrence, I couldn't help noticing that the aisles of the grocery section were being transformed - from a confetti of colors of the labels of the national brands to the stark white of generic "Great Value" goods.

And while "Great Value" is technically a store brand and not a true generic, my sense is it can be considered the equivalent. Most stores attempt to instill a sense of quality in their store brands - to suggest that they are "just as good" as the national brand - while Walmart's store brand is merely cheaper.

What struck me afterward is the remembrance of other times in recent history when generic products appeared on supermarket shelves. During the recession of the early 1980's and the savings-and-loan meltdown of the early 1990's. I didn't notice them in the way of the dot-com crash at the turn of the century - though the effects weren't as widespread. But they have now reappeared, in the wake of the housing crisis.

I don't expect this means that the appearance of generic goods could be used as an indicator that the economy is about to go south. They generally come after the fall, when people are dealing with financial difficulties and can no longer afford the quality of goods they normally prefer.

But on the other hand, they have tended to disappear as the economy has recovered, so their appearance provides little information of value, besides what is already known and felt by a great many people, but their gradual disappearance might be a genuine sign of economic recovery, more reliable than the statistics and indicators trotted out in a desperate attempt to improve the disposition of taxpayers and voters.

In the end, I'm not certain whether the matter that's on my mind is one of economics, or the old price-versus-quality debate of marketing. Perhaps it's just an observation, the meaning of which will become clearer (or dissipate entirely) with further reflection.

Wearing In

I caught a video clip in which a designer discussed the idea of items that "wear in" instead of wearing out. It's an interesting notion, and one that applies readily to physical objects: the pair of shoes that get more comfortable over time as the leather softens and molds to the wearer's feet, or the shirt that feels most comfortable when it's on the verge of being threadbare.

His premise was that objects should become more useful over time rather than less, and that age and use create patterns of wear that make them more comfortable or more efficient. It seemed a sophisticated way of making a common complaint: that things aren't made as well as they used to be, and that the demand for continual profit leads manufacturers to make things that will become useless over time to encourage consumers to replace them. Planned obsolescence as a source of revenue.

I considered this notion for a while, especially in terms of the digital media - not so much the objects we use, computers and cell-phones with a practical life of several months before a newer and better model makes comes along and the software companies no longer develop for them or support their "old" software. It seems that the digital media is inherently compelled to be new and different instead of old and familiar, and my sense is that's not always desirable.

There is a certain level of fascination with the new, but there is also a certain level of frustration. I knew how to use the last version of the software, but this upgraded version is completely different and I can't figure out how to make it do what I want (not to mention that it is loaded with features I will never use, consumes so much memory and processor I now need to upgrade my computer, etc.). I knew how to find things before the company redesigned its web site.

And this leads to another problem: what's the point in learning to use some computer software or a Web site if it's going to change in six months? My sense is software designers themselves are to blame for one of the biggest headaches they complain about: user stupidity. Nobody reads manuals anymore; nobody refers to the on-screen help; nobody invests much time in learning to use the software, but simply calls the support number with complaints that the software or Web site is badly designed.

And in this instance, I have to side with the users: it's too much to expect a user to learn and remember a sequence of actions to accomplish the same task when that sequence changes with every version upgrade or site redesign. The software wasn't designed to "wear in" - and what it wore out, ultimately, was the patience and tolerance of users.

Communism and Customer Service

While reading an article unrelated to my typical fields of study, I came across an interesting phrase:

"Communism's demise was inevitable because it offered respect to the community but never to the individual, and so drained itself of vitality and spirit, one citizen at a time."

It's not my intention to inject a political element, merely to cite the source of a rumination: that the same can be said of businesses in regard to customer service. Change a few words in that quote, thus:

A company's demise becomes inevitable if it offers respect to the market but never to the individual customer, and so it will drain itself of revenue, one customer at a time.

I think that summarizes, rather aptly, a misgiving I have had about the way in which the notion of "user experience" is often put into action. The focus is on a generalized conception of all consumers, as a group, while dismissing every individual that comprises that group.

And taken to extremes, it becomes a sort of bigoted arrogance to declare to a customer that you know what he, as an individual, expects in the way of service because you believe him to be a member of a homogeneous group of people who all think and act the same way. And there is no quicker way to alienate a customer than to declare "I know what you need," and then rattle off a list of things that he, as an individual, doesn't even want.

Such practice is contrary to the most basic tenets of customer service: to determine what a customer wants or needs, and then deliver it. Granted, it's marginally better than the introspective approach of assuming the company knows what a customer wants, or ought to want, without even asking - but it still falls short of the mark of providing excellent service to each individual customer.

Artistry vs. Design

Someone new to UX recently asked me how I come up with "ideas for cool designs." My response was simply that it's a philosophical question, something we should discuss when we had more time, but that fundamentally, "cool" is not the foremost concern of design, and that he should focus on making things useful, and leave aesthetics to the young PhotoShop jockeys who will turn his wireframes into comps. Then, we went along to our respective meetings, and it will probably be a long while before we can find an hour to sit and chat about it.

But it got me to thinking: The notion of "design" is too often regarded as artistry instead of craftsmanship. "Design" implies color and shape, the arrangement of elements into a composition that is visually compelling. I can't dismiss the importance of this - people gravitate toward beautiful things, and tend to want to interact with them, and instilling in users the desire to touch an object or click around a Web interface is important.

But it's also quite short-lived. Aesthetic appeal causes the user to become interested and interact with an object. But there, the contribution of the artist ends. If interacting with an artful object serves no purpose, or if it is difficult to achieve the goal of interaction because of compromises made to make it attractive, the object fails its purpose.

The initiation of interaction is where "artistry" becomes less important than "design," where a object that merely looks cool is found, in fact, to be entirely useless. The user abandons it, eventually if not immediately, and is left with a sense of disappointment or frustration. Whatever the case, he's lost interest in interacting with it, and is a little less likely to be fooled by appearances in future. In that sense, artistry and design can be in conflict.

My sense is that, usually, design is more important than artistry. Design delivers the value that people seek in objects, and I think that it is this delivery of value that creates the emotional connection between people and things. It is the reason why we are fond of the "old" - why we won't part with the jacket that is no longer in fashion, and is a bit the worse for wear, but which is comfortable and warm.

And my sense is that perception of this value is a matter of experience - why the old carpenter won't part with the trusty hammer he's been using since he was an apprentice, while the young apprentice brings to the job a chrome-plated model with a molded plastic handle stamped with the brand of whatever manufacturer has done the most advertising lately.

Ideally, both art and design can be accommodated. The two aren't always at odds, and are not necessarily in conflict. If the "cool" hammer is also well-designed, then it's likely that thirty years hence, it will still hang from the belt of today's "new" apprentice.

But where the two are in conflict, in an instance in which making something "look cool" means sacrificing what "works well," a compromise must be made. I would value design over artistry, but that's a decision made from my own perspective and based on my own experience. Ask someone else, and you'll get a different answer.

Budgeting as a Stifling Ritual

I don't often meditate on accounting. That's not to say it's unimportant - ultimately, profit and loss is what drives a commercial enterprise - merely that it's entirely uninteresting and seldom the source of much innovation in a business. It's a backward-facing practice that assesses the financial impact of what has already been done, and which only looks up from the ledger once a year to consider what might be done in future.

And that's precisely what's got me to thinking: bean-counters have a great deal of control over a firm's activities by virtue of having control over the resources necessary to do anything. Even when you set aside the potential problems (which are many) that arise when the bean-counters overstep the bounds of their authority (which even when seldom is too often), there is an inherent problem with the infrequency with which financial decisions are made by a firm.

The core of the matter comes from budgeting schedules: the fact that, in many organizations, budget is allocated only once a year (there is an annual budgeting cycle) enables better control of financial resources, but prevents firms from being innovative by preventing ideas from coming to fruition in a timely manner. In effect, the budgeting process is one of the major factors (if not the most significant among them) that firms become stodgy, slow, and unable to outmaneuver the competition.

In plainer terms, budgeting puts good ideas on a shelf, for no other reason than to make life easier on the individuals who prepare the budget. If someone has a great idea in February that will generate $1 million per month for the firm, it has to wait until October to be considered, and the following January (at least) to be implemented. Which means the company forfeits $10,000,000 for not having implemented the idea sooner.

Why? Because October is the time when we do planning and budgeting for the next fiscal year. Why? Because we do. I can conceive of no other reason.

Perhaps it's a throwback to the time of ink and paper, where gathering up information about a firm's financial status and ongoing operations required months of effort and was terribly expensive, such that it made sense to do it only once a year.

Given computerized accounting systems, which make information available almost instantaneously, that excuse is no longer valid. A firm of any size should be able, within a very short amount of time, to assess its operations and make decisions. The present fashion of "dashboards" that provide a real-time view of data should facilitate this - though my sense is that most companies see dashboards at a fad and do not use the information to drive decision-making, but instead, give them an occasional glance. (Which could just be an effect of the metaphor - a dashboard speedometer doesn't tell you which way to steer.)