| Solutions |
|
 |
 |
|
|
 |
||
Tetra Pak: Using the intranet to drive productivity and efficiency Contents
Task performance Indicator @ Tetra Pak Fix the basics Managing processes, not projects Fixing those confusing menus and links Designs A, B and C Helpful search results Content quality is as basic as it gets Smelly content must go It's about time Top task management at Tetra PakTetra Pak has taken the concepts of top task management for an intranet further than any other organization I have worked with. I asked their e-Communications Director, Gabriel Olsson, why this might be. "We are both the sons of farmers," was his reply. The point he was making was that, as a farmer's son I should understand that task management is like farming. Perhaps because farmers know about the need to save time on tasks, and have been doing so longer than many industries. It is about getting the basics right and being willing to do the boring but necessary stuff (like content review) on an ongoing basis. Rolling your sleeves up, focusing on what is important to your employees, not just what is cool and fun for the web team.In 2006 I was asked to help Tetra Pak to review its intranet strategy. Tetra Pak followed the intranet management models that all organizations I have worked with have followed. They had initially seen it as a technology issue and thought the key challenge was choosing the right technology with little attention being paid to the needs and habits of the employees. This approach had limited success. The other key strategic belief was that design and publishing should be distributed as much as possible. Basically, whoever wanted an intranet site and whoever wanted to publish got some training. They never kept to the use of one 'right technology' as there was much freedom in the company culture and limited global planning and management. This approach, while having some benefits, had lead to the classic case of the largely unmanaged and unmanageable intranet. The Tetra Pak intranet grew, and grew fast. Nobody was quite sure how many pages there were but rough estimates of 500,000 were given. Employees were frustrated. Here are some quotes from staff:
The intranet's number one objective is to make staff more productive. We decided to do a top task analysis and over 800 staff (out of 20,000) voted on their top tasks. The top three task areas were:
Task Performance Indicator @ Tetra PakThe next step was to measure the completion of the top tasks. In 2008 we carried out extensive task measurements. The results showed that there was a lot of work to be done.IMAGE Of the tasks tested only 61% were completed successfully. 32% of participants failed, while 7% had disasters. (They got an answer they believed to be correct but which was in fact wrong.) The logic of task management is that if you are managing tasks, not content or technology, then you should measure based on task success, and that's exactly what Tetra Pak has done. The management conversation is not about getting a new content management system or search engine or adding more intranet sites or content. It is about improving task success rates, reducing disaster rates and improving completion times. The improvement of the task or an increase in employee productivity is the final objective, and to this end the people involved in running the intranet are called "task drivers". They drive the efficiency of the tasks and are measured on whether or not they have achieved sufficient efficiency. As a result of measuring a wide range of tasks it became clear that there were three underlying factors that were affecting task performance regardless of the type of task:
Fix the basicsIt would have been very easy for Tetra Pak to try to solve its problems by initiating some major projects to install new software and/or carry out a redesign of the website. It's amazing how many organizations take this knee-jerk tactical approach. The IT department gets involved and wants to solve the problem by buying new technology. Communications gets involved and wants to solve the problem by changing the graphical design.At Tetra Pak, both Communications and IT reached a consensus. Firstly, the challenge was complex and could only be solved if both departments worked very closely together. There was a deliberate and active attempt by each department to understand each other. That was just the start of the collaborative thinking. While formal responsibility for the intranet lay within a section of Corporate Communications called e-Communications, it was quickly understood that making the intranet better would require cross-organizational support and commitment from the business. Creating clear menus and links, improving the quality of search results and getting rid of out-of-date content were challenges that needed to be addressed by every single person involved in publishing to the intranet. It didn't matter if they worked for Communications, Finance, or Products; or if they worked in Sweden, America or Italy. Success would be achieved by an organizational effort. So, a lot of time was spent winning people over to the top tasks approach. It was hard work. The motto for the new approach was: fix the basics. Managing processes, not projectsFrom all the data we have analyzed about why people are dissatisfied with their ability to complete tasks on websites, "confusing menus and links" is significantly ahead as the most important issue. "Poor search results" is also very significant, but confusing menus and links is the top issue for both intranets and public websites. It is the number one issue for government and commercial websites. It is the number one issue in the United States, Canada, UK, Sweden and Norway. Basically, everywhere we test it is the number one issue.Tetra Pak embraced the continuous improvement philosophy based on employee behavioral evidence. "Facts, not opinion" became a mantra. And that's the way the Web is going; towards making decisions based on behavior data. The continuous improvement model eliminates the need for big redesign projects. The project culture has often been the enemy of quality. The project itself becomes the obsession; the redesign, the graphics, the new tool. Management bonuses become contingent on the project; "implement a CMS", or "redesign the website", for example. There's a deadline, and that deadline must be met. Something must be launched. It doesn't really matter what-just launch something so that a box can be ticked and management higher-ups can be told: "Hey, look at us, we're doing stuff, we're delivering stuff. We deserve a bonus". Fixing those confusing menus and linksNavigation design is a complex activity. It takes lots of thought and testing. It is the information architecture-the foundations-of your website. Get it right and everything else becomes so much easier. Get it wrong and the more content and tools you add, the more confusing your website will become. In other words, the more you add-the more work you do-the worse your website will become.In one navigation test we did with Tetra Pak employees, the content owners said that for a particular task, employees would know to click Link A, because that's where this content had always been kept and everybody knew that. So we tested with 15 typical employees. Only one clicked Link A. In implementing the continuous improvement model, Tetra Pak developed the following guiding principles for the navigation design:
Design AHome | About Tetra Pak | About You | Products & Sales | Technical | Processes | CollaborationNext, we did a simplified version of the task performance indicator. We took the top task questions. Questions like the following:
How many packages per hour can the TBA19 filling machine produce?
We agreed on where we expected the employees to click, gave these top task questions to a representative sample of employees, showed them Design A and asked them:
Where would you click first to complete this task?
The data quickly illustrated that there were flaws with our initial design, with Design A having only a 60% success rate. We measured success based on whether people clicked the links we expected them to. Have a look at Design A again:
Home | About Tetra Pak | About You | Products & Sales | Technical | Processes | Collaboration
Design A contains classic flaws of navigation design. What does "Technical" mean to an engineering organization like Tetra Pak? We might as well have a classification called "Tools" or "Infinity & Beyond". It only caused confusion and annoyance. What has been noticed through much testing is that when you present people with classifications that they find illogical or too all-encompassing, they begin to lose confidence in your design. Some specific things we discovered from this testing included:
Design BHome | About Me | About Products, Marketing & Selling | About Technical | About Procedures & Rules | About Communities | About Tetra PakDesign B resulted in a 70% success rate, a 10% improvement. However, there was strong feedback that employees did not like having "About" before each class. Finding people is the top task on most intranets and we had noticed in this testing and in the overall task performance indicator tests that failure rates for this particular task were quite high. Then someone had a bright idea: Why not have a Find People classification? Under it we could have the find people search, the organization charts, communities and various other collaborative activities. So, we brought out the next design. Design CTetra Pak Home | Products | Find People | About Me | About Tetra PakThe test results for Design C had a success rate of over 90%. Of course, this was just the start of the process. Every major level of the classification was subsequently tested in order to improve task completion success rates. Helpful search resultsOne of the most important things Tetra Pak did was to make people responsible for the task success. It was no longer enough to make an employee responsible for the content or the technology or the search engine. The search professional appointed had to be responsible for the findability and successful completion of the top tasks.Tetra Pak carried out the Search Performance Indicator process. To sum up:
Tetra Pak has a search long neck on its intranet: A small number of search terms get a huge number of searches. "The situation for search was bad," Gabriel Olsson stated. "When we started to measure, only 28% of the searches were successful." How was that score calculated? Let's say you're testing 300 search terms. The maximum score you could get would be 300 X 42 = 12,600. In other words, the correct page was first for each of the 300 search terms. So, a 28% score means that Tetra Pak got 3,528 out of a possible 12,600. Why was search failing at Tetra Pak? Most organizations would immediately blame the search engine and a project would be set in place to buy a new one. The search engine that Tetra Pak was using did indeed have limitations but, as Gabriel pointed out, "the search application is often blamed as the only reason why search fails, and that is not the case. There is much one can fix and improve before the need for a new search engine is obvious." A more important factor was the quality of the content. So, the next step was a rigorous review of the content environment with a particular focus on the first page of search results for each search term. What was found was:
Manual recommendations (best bets) were also used. This involved the search professional introducing into the search results a recommendation of what they thought was the best result for a particular search term. Within 6 months of this effort, the success rate had moved from 28% to 57%. No new search engine, no major expenses. Just a rolling up of the sleeves and a fixing of the basics. It's as old as computing-garbage in, garbage out. Too many web teams want to build things that will help them land on the moon, when they should be sweeping the floor and washing the dishes. Great web teams are not afraid to do the boring and repetitive; to grind out the incremental improvements. It's sexy and exciting to do the big redesign, but there is nearly always far more value-at far less cost-to be achieved by fixing the basics and adopting a day-by-day focus on the essentials.
Content creators are measured not based on the use of the content but rather on its production.
Content quality is as basic as it getsThe first time I met Mats Johansson was at a workshop I was giving in Copenhagen, circa 2002. At that time Mats was responsible for the technical services and support section of the Tetra Pak intranet.Mats believed in content quality when very few others did. It's strange that even though the Web runs on content, so few people actually care about content quality. It's as if any old content will do. Which, of course, it won't. If you want a cast iron, absolute, definitively guaranteed way to lose your employees' trust, give them inaccurate or out-of-date content. It works every single time. When Mats worked for Tetra Pak, he would gather his intranet publishers every couple of months and go through the content they were responsible for. As part of the review process, he would show them the following slide. IMAGE He explained to them that if they, the publisher, spent only 10 minutes on the content because they didn't have time to do it right, they forced the visitor (the employee) to spend 60 minutes because it was harder to find the content, harder to understand it, harder to be sure that it was correct and trustworthy. If there were five people who needed this particular piece of content for their jobs then the overall cost to Tetra Pak was 310 minutes. And who are these visitors? They are Tetra Pak service engineers and support staff. They do important and valuable work. Their time is precious. Poor putting-up is unproductive, big time. Then, Mats showed the publishers the next scenario. IMAGE Let's say the publisher now spends 60 minutes on the content. Well written, clear, simple, bullet points. Good metadata so it's easy to find. Well, now the visitor only needs to spend 10 minutes finding and understanding it. The total cost of time to Tetra Pak is 110 minutes, which is a lot less than the 310 minutes in the first scenario. And think about the potential time savings if you had 1,000 visitors! Sounds like a no brainer that the publisher should spend more time creating quality content. Not so. Not the way modern organizations are managed. You see, the publisher's time is measured. They work in Support or Finance or HR. The pressure is on for them to spend as little time as possible. They are not measured based on the use of the content but rather on its production. The visitors' time is invisible. Real it may be but it is spread throughout the organization. Nobody is responsible for it. It is not measured and we know that if it is not measured, it is not managed and if it is not managed then it cannot be important because the important things are managed. There are other factors at work here. Often the content that is produced by a particular unit or department is for consumption by people outside that department/unit. The people working in the department don't need to read the content after they've published it because they are experts, and even if they are not they can find an expert close by. So, they don't see the impact that poor quality content has because they don't work with or go to lunch with the people it affects. They may genuinely think the content is simple to read. They may think it is logically organized, and it may well be excellently organized from a departmental point of view. But the classification and organizational structure that allows people to work efficiently as a team is rarely the classification that makes sense to the customer looking in. Smelly content must goOut-of-date content may not smell, but it sure does stink. It is one of the biggest website management problems. Nobody would leave a bowl of fruit on a desk for months but we have no problem leaving old content on websites for years. Why is that? It's because the predominant culture within organizations is one of launch and leave. Once the content is put up that's the job done as far as the publisher is concerned.But unlike print content, web content never goes away. Once you publish it on your website it remains findable by search engines and remains embedded in the navigation and links. Out-of-date content was not as big an issue within Tetra Pak as I have found it to be in many organizations (most intranets are glorified dumps), but it was still a pretty important factor in the poor ratings of the intranet by employees. Elaine See, Intranet Program Manager at Tetra Pak, used to sell online investor-relations programs to Japanese corporations. She learned a lot of quality management techniques as she met the needs of some of the most demanding customers in the world. As head of task management for the Tetra Pak intranet, she was keenly aware of the need to make sure quality standards were achieved. Supported by the content quality task driver, Elaine began to put the following strategy in place. An initial audit was done of the whole intranet. It was estimated that there were around 500,000 pages. "No wonder you couldn't find what you were looking for," Elaine stated. So, one of the first steps was to remove content published before 2006 and to set a policy in place that no more than 2 years of content be kept on the intranet. This sounds quite draconian but it works surprisingly well. In reality, there is very little truly old content that is of much value and where there is, exceptions can be made. By following this policy the amount of pages was halved. This was just a first step. The critical part was getting a review process in place. "On a local level there are a few units that do have regular reviews, but we need to make sure that we have a global review process to follow," Elaine stated. So Tetra Pak hired temporary staff to review every single piece of content and contact every single person who had ever published. The publishers were informed that this would not be a once-off process and that there would be a minimum of a 6-monthly review cycle. What were the review criteria? "In essence, quality content must support a task", Elaine stated. "Key questions to ask are:
It's about timeTetra Pak has started to examine how task efficiency affects productivity. In just one area it has found that improving five tasks results in a time saving corresponding to more than €125,000 per year. As another example, in the Human Resource section for Sweden they needed one less person for phone support after they introduced the task-based approach. Feedback from employees summed up the new world:
Email Customer Carewords The above is an excerpt from Gerry McGovern’s latest book, The Stranger’s Long Neck France: Buy from Amazon Germany: Buy from Amazon Japan: Buy from Amazon |
Customer Carewords is a unique scoring technique enabling customers to tell you in precise, statistical terms their top tasks when they arrive at your website. More |
Home | About | Solutions | Customers | Contact