August 2013


From Managing Inputs to Managing Outcomes

CIDMIconNewsletter Gerry McGovern, Customer Carewords

The Increasing Value of Quality Content

We are in the middle of a value revolution. Value is moving away from physical things and toward interfaces. It is moving toward content, whether that content be in pictures or words. Content is increasingly our first contact—our first interaction—with an organization. When we want to do something, we go to the Internet and the Internet is a world made up of content.

Most organizations do not value content. They see it as a cost. The value is in the product or service. This attitude can be even more pronounced for technical and support content. This is stuff customers get after they buy the product, the thinking goes. The value from the customer has already been achieved. The sale has been made, so let’s manage after-sales costs as tightly as possible.

A localization manager was once asked what the job entailed. “My job is to get people to create unreadable English so that it is easy to translate into unreadable German.” Localization has, in many situations, become a race to the bottom with a relentless focus on squeezing out costs wherever possible. Rarely is there a conversation about how much value this localized content can create.

Our web sites are the goldmines of the Internet age, yet we are managing them like coalmines. Often, one word can make all the difference. A large B2B company we’ve worked with had a link called “Find a dealer.” We convinced them to change it to “Find a reseller.” Leads to their resellers doubled. We experimented more and changed it to “Find a local reseller.” This led to another 50 percent increase in leads.

 The Old Metrics for Content are Broken

An organization we dealt with was beginning to test its technical information with its core engineer audience. One question was like this: “Can you do X with feature Y?” The answer was no. The engineers didn’t have a problem finding the page the answer was on but every single engineer we tested said the answer was yes, you could. Why? Because the organization thought that it was so important to stress that feature Y could not do X that they put it in a “Special Note.” They put this special note in a nice colorful box and placed it in the center of the page.

Everyone ignored the special note in the special box with the special color because it looked like an ad. It was desperately trying to get the engineers’ attention so they automatically thought that it must be useless information. So, the techniques the writers were using to give particular emphasis to things were in fact de-emphasizing those things. And the writers didn’t know this because they were not getting proper feedback on the use of the content.

After content is published, it goes into a black hole as far as most organizations are concerned. There are no useful metrics that measure success. Volume-based metrics often measure failure. For example, Microsoft Excel used to think their function pages were very popular. But that’s because these function pages had titles containing words such as “Imsum, Areas, Print” so many people who wanted to know how to print a page or sum two numbers ended up on function pages. And they weren’t at all happy. Microsoft Excel ended up either deleting or hiding from search half of their pages. The result? Their first major increase in customer satisfaction in years.

Telenor, a major Scandinavian telecom company, deleted almost 90 percent of their pages. Conversions went up by 100 percent. Support requests went down by 35 percent. The city of Liverpool deleted 80 percent of their pages. Support calls were reduced, and online reporting went up by 400 percent.

The traditional publishing industry has encouraged a culture of production, of volume. In the print publishing world, it is easier and cheaper to publish 100 pages in one book than 20 separate pieces of two pages each. In print, there are economies of scale and a specific business case, but online it’s often the opposite. Instead of economies of scale, we have economies of attention.

Bad Content Kills Good Content: Top Tasks Versus Tiny Tasks

The web is not print. It’s just not. The web has different rules. The web has a different business case. The web is much more about getting the exact piece of accurate information in the most usable form as quickly as possible. Amazon would not have lasted for long if all it offered was a big PDF of its book catalog that was two months out of date. (PDFs are most definitely NOT web content. They are print content masquerading as web content.)

When it comes to the web we need to focus on outcomes, not inputs. Did the customers quickly find what they needed? Did they easily understand it? Were they able to act on it? We can focus all we want on production and publication but if the information is not findable and useful, then it has no value. In fact, it may have negative value if it gets in the way of more useful content. (IMSUM function pages had negative value when people were trying to find out how to sum/add a number.)

It’s hard to make information useful on the web if all you’re dealing with are large PDFs that were designed for print or endless HTML pages that have been produced by essentially copying and pasting print content. It’s not very useful to say to a customer “The answer is on the web site somewhere.”

Historically, the web has been driven by a Cult of Volume. There was an assumption that if we just write about everything, the magical search engine will sort it all out. I once spoke to a Microsoft executive who estimated that the Microsoft public web sites had about 15 million pages and that about 4 million of them had never been accessed. That’s roughly the equivalent of the population of Ireland in pages that have never been accessed. Why was all the time and effort put into publishing these pages?

It’s not simply that publishing content nobody wants is a waste of time and money. It’s also that the more content you have, the harder it is to find the right content. One of the reasons is word overlap. If you want to know how to print a page in Excel and you search the Microsoft web site for “print page,” then the search engine will bring you back pages with the word “print” in them, and one of them used to be a result called “Print Function.” (The Print Function is a mathematical formula but it has the word “print” in its title.)

This is what we call a ‘Top Task versus Tiny Task’ problem. The top task is to know how to print a page. The tiny task is to know how a print function works. There are many, many tiny tasks and over time they get in the way of the top tasks. They choke the search results and clutter the navigation paths.

Worse, when a tiny task goes to bed at night it dreams of being a top task. When it wakes up in the morning, it’s gung ho. It wants to be a top task, and it goes to the web team pleading and trying to persuade them. We have found that often the web team’s week is wasted dealing with tiny tasks. They are nibbled to death by tiny tasks.

Top Tasks Management

Customers are incredibly impatient today. If we don’t deliver the information they need in as fast and simple a way as possible, they leave. Achieving that degree of speed and simplicity requires extensive and ongoing design and management of information. How to do that considering we are producing so much information?

The answer: Produce much less information but manage it much more.

Top tasks management helps you do that. It involves clearly identifying the top tasks of customers and the tiny tasks. Then, measure how well the top tasks are performing, show how the tiny tasks are impeding the performance of the top tasks, and remove them.

We have worked extensively with Cisco since 2009 to identify customer top tasks. We identified about 80 major tasks for the web site. When we got customers to vote on these tasks, the top 3 tasks got as much of the vote as the bottom 43. This is a classic pattern we have found in over 400 similar studies in more than 15 countries for some of the world’s largest organizations.

The number one top task for Cisco customers was “Download software, firmware, updates.” Cisco then created typical examples of this task such as: “Download the latest firmware for the RV042 router.” We got a representative sample of engineers to do this task and measured them:

  1. Success rate: How many engineers successfully completed the task?
  2. Disaster rate: How many thought they had downloaded the right firmware version but it was the wrong one?
  3. Completion time: How long did it take to download the firmware?

Cisco is now focused on outcomes, not inputs. The latest firmware being published on the web site is no longer a measure of success. Success is now measured by how quick and easy it is for customers to find and download the firmware. The result? In 2010, it took an average of 280 seconds and 15 steps to download a typical piece of software. By 2012, this had dropped to less than 100 seconds and 7 steps. And Cisco works every day to make things easier and faster for their customers. It benchmarks its top tasks on a quarterly basis and plans its web site activities around improving task performance.

Support is the New Sales

Over the years, we have helped analyze customer behavior for organizations such as Microsoft, Cisco, IBM, VMware, and so on. One key trend we have noticed is that potential customers have become much more demanding in the type of information they seek. Increasingly, they are looking for technical information that organizations assume is only required by people who have actually bought the product. What do these potential customers want?

  1. Installation, configuration, set-up deployment
  2. Using the products (get started, how-tos, tutorials)
  3. Maintenance, operation
  4. Code samples
  5. Troubleshooting (bug fixes, diagnostics, guides)
  6. System design
  7. Technical communities
  8. Training

It’s like the potential customer is coming to your web site and wants to take your products for a test drive. They ask questions like

  • What’s the installation process like?
  • What’s it like to troubleshoot?
  • What do existing customers say about the product on the communities?
  • What’s it like to maintain so I can get a genuine total cost of ownership?

Based on our analysis of potential customer behavior we can say that anywhere from 30-50 percent of the tasks they want to complete when they come to your web site are not addressed by your traditional marketing material. Support is the new sales. Technical documentation is the new sales.

We have also noticed another pattern. Customers don’t think in solutions, at least not when they’re on the web. They overwhelmingly think in products and in the context of the very specific problems they have. When they get to a web site, the first thing they often want to do is get to the product homepage. They want to get to Product X homepage and then look to troubleshoot Product X, to find out about installation for Product X, to see how easy Product X is to maintain.

But that’s not how Product X is organized on most web sites today. How is Product X organized?

  • We have a marketing homepage.
  • We have a support homepage.
  • We have a community homepage.
  • We have a technical documentation homepage.

This ‘silofication’ of product information confuses and frustrates customers. It also loses sales and reduces customer satisfaction. With one company we worked with, adding installation links on its product marketing pages had the single biggest impact on increasing sales conversions. All the other marketing banners and content paled in relation to the power of the installation guide to help sell the product. 

From Coalmines to Goldmines

Detailed technical information has never been more in demand. Customers want to know more in order to make better decisions. But the information must be easy to find and easy to understand.

Most organizations do not invest enough in their information because they only measure inputs. They only measure the cost of producing the information. They treat information like it was coal—a low grade commodity to be dealt with in bulk. They do not measure the real value that information can create.

Until we had the web, it was very difficult to measure the value of information. But the web is an amazing laboratory where we can observe human interactions with information. We can now measure outcomes. Once we start to measure outcomes, we will realize what an amazing goldmine we are sitting on. CIDMIconNewsletter

About the Author


Gerry McGovern
Customer Carewords

Gerry has published five books about the need for customer-centricity in the online world. His latest, The Stranger’s Long Neck, focuses on how to identify and measure the top tasks of customers when they are online. He has spoken and consulted on website management in 35 countries. His company, Customer Carewords, specializes in helping large organizations (such as Microsoft, Cisco and VMware) make their online presence more successful by improving the performance of the top tasks of customers.