Extra! Extra!

In a recent FastCo.Design article about brainstorming, Cliff Kuang discusses how criticism during group brainstorming can actually improve the flow of ideas. I agree – as long as the criticism is productive and generative. Here are 10 methods for ensuring your criticism makes ideas flow during the next design or wireframe review you attend:

1. Establish Goals First. Make sure there’s consensus around what the goal of the design, interaction, page, or product is before starting to review it.

2. Ask Questions. Ask questions to understand the intentions behind decisions. Give the designer the benefit of the doubt by asking questions like, “Did you think about…?” and “What would happen if…?”

3. Discuss only after Understanding. Ask clarifying questions readily until you feel you could explain the idea to someone else. Wait until you understand the full idea and concept before discussing changes. When you do discuss changes, build off someone else’s idea instead of going in a complete 180 direction.

4. Use Sticky Notes to Comment. Actually putting sticky notes on the design gives everyone in the group an equal opportunity to offer insight. Sticky notes are also removable and movable!

5. Move through the Design with Scenarios. Returning to scenarios brings the user to the forefront and asks, “What would the user do?”

6. Explore the Pros and Cons of Alternative Approaches. Talking about alternatives helps to eliminate assumptions and ensure that each member of the design team holds the same priorities.

7. Project the Design on a Whiteboard to Allow for Writing and Sketching in Context. This activity gets everyone up and moving during a critique and captures all of the feedback in context.

8. Compare the Design against Nielsen’s Heuristics. Heurstics serve as a great reminder checklist of basic principles of usability.

9. Use Metaphors to Express Criticism and Ideas. Metaphors can be easier to talk about objectively than the brainchild of the design team. They also focus the attention on the broader sentiment of the design.

10. Don’t Forget to be Positive! Constructive feedback is a as valuable as criticism. Plus, it shows your respect for the design team’s ideas and work.

Market research is helpful to User Experience Design by extracting behavior-centric information to inform the online strategy. The role of the User Experience Designer is to get to know not only who the customers are on the web, but also what they do, and how they do it. Those insights can begin to shape from the clients Marketing Department. Here are a few examples where marketing categories containing research about customers can be easily translated into practical insights that are helpful to UX:

Marketing Category - Branding Guidelines - Insights for UX -What messaging and style will be appealing? What does the training material explains about how to interact with customers and troubleshoot common problems.

Marketing Category - Market assessment and target demographics - Insights for UX - The types of workflows, user tasks and practices seen in the field.

Marketing Category - Promotion content materials - Insights for UX- What incentives make people buy the products and create loyal customers. What opportunities are there for up selling?

Marketing Category – Marketing Roadmap - Insights for UX - Which audience/demographic they want to grow and which customers and users are most important? What are their competitor websites and what they like & what they don’t like about them. Which products or aspects of the company will grow and which ones will go away.

Designing and moderating usability tests requires the specialized skills of understanding how people’s brains work. Even more so, interpreting the results from usability testing requires the specialized skill of switching between meticulous details of interaction and broad strokes of experiences across participants.

Here are 5 tips to avoiding pitfalls when interpreting usability test sessions:

1. Suspend judgment. While moderating usability tests, it’s easy for me to get caught up in the moment. I feel so close and empathetic towards each user that it’s easy to take what one person says and run with it. However, I have to remind myself that I’m not designing the system for one user or even one type of user most of the time. I force myself to suspend judgment until I can go back and review the recordings of the sessions. I recommend both video capture of the participant’s face as well as of the computer screen.
2. Look across users. As a moderator and human being, I find I have biases in terms of what I find interesting in a usability session or what I take notice of during a usability test. Having a partner in crime watching the usability test is the best insurance that our team will get unbiased results. Another good technique I’ve learned for eliminating my biases is lining each task of the usability test in a column on a spreadsheet. I go through each participant and put details from each participant’s experience in the rows corresponding to the column tasks. This way, all the participant’s details (including completion rate, portions of the task they struggled with, quotes, etc.) are lined up and easy to scan for patterns.
3. Listen, no really listen, to participants. If it is clear to us that we missed the mark on an interaction, it is very helpful to play/rewind/play/rewind the video tapes to listen to how participants talk about the interaction. It’s so valuable to see what participants do, but it’s equally valuable to understand their mental models and how they see the feature through the language they use to describe it. Recording quotes from each participant can offer some surprising insights into how to course-correct a feature to match people’s mental models.
4. Know the limitations of testing. I’m an advocate of usability testing, but I know there are limitations for how realistic it can be. Sometimes our prototype doesn’t provide the same flexibility the real system would. Sometimes we are testing something new that might change people’s workflows over time. When looking at the results of the usability test, I have to remind myself what we were testing and what we were NOT testing. This helps remove some of the confusion and noise of the data.
5. Understand trade-offs. There’s no such thing as a free lunch. As a user experience designer, there’s no such thing as a perfect design. When making decisions about what needs to change from a usability testing session, I have to remember that each feature does not exist in a vacuum. Changing the details of one feature could negatively impact another feature. Before making recommendations, I consider all the implications of changing something in the ecosystem. If there is time, iterative testing is a great way to really understand these trade-offs.

I’m enthusiastic. I’m optimistic. I’m really looking forward to see what 2012 will do for the web analytics industry.

Here are some predictions for web analytics trends in the year ahead.

1. IBM and Adobe will succeed where many have failed:  Acquisitions are tough in the web analytics space.  The space innovates quickly and has re-invented itself several times in the last decade.  Microsoft, Yahoo, and Digital River all made expensive acquisitions in the space and then failed to innovate losing market share into obscurity.  In 2012, Adobe will bring its expertise in interface development to the web analytics space and IBM will bring true enterprise analytics to the web analytics space taking the industry in new directions.
2. Exclude (Windows|Macintosh|Linux|iPhone) will grow quickly:  Create a segment that excludes all traffic from the top 4 operating systems and you will find a small group of visitors that probably grew by over 200% in the last year.  Visitors from platforms such as Android, Google TV, Playstaion, Nook, Nintendo, BlackBerry, etc are growing fast. While they only account for less than 7% of the traffic today, at this rate of growth they will be a significant audience by the end of 2012.
3. Audience scoring: Web analytics solutions are great at reporting on audiences based on their source, conversion, searches, etc.  Most analytics tools fail on grading audiences based on their behaviors in between the first visit and the last.  The fact is, some audiences are more valuable (i.e. more qualified) than others and web analytics tools are probably the most accurate method for screening these visitors.
4. Heatmaps and session recordings become main stream: When multivariate testing had explosive growth, each of the major analytics providers purchased and integrated multivariate solutions.  In 2011 heatmaps and session recordings grew quickly as a instrument in the web analytics tool kit.  In 2012, we hope the major analytics providers will purchase or develop their own heatmap solutions.
5. Integrated channel analytics becomes real:  With all major analytics customers focusing on a complex digital channel including YouTube channels, FaceBook applications, eMail marketing, 3^rd party forums, etc the web analytics vendors have done a poor job offering a satisfying solution for this need.  In 2012 we are hopeful that the analytics vendors will find creative solutions for reporting on a complete multi-channel lifecycle.
6. Tracking user interface events (HTML5/AJAX/JS):  HTML5 is expanding quickly.  AJAX usage continues to grow.  More functionality is coming to the interface every month.  The current tools for tracking interface events are poor with insufficient debuggers.  In 2012 we want to see better tools for creating tracking code and any thing the vendors can do to help with debugging will be appreciated.
7. Analytics containers will continue to grow in popularity:  Tag containers help ease the implementation and updating of analytics tags across websites.  With Tag containers, the analysts can make updates to the analytics without using engineers or IT staff.  With a higher customization of tracking tags in 2012, more organizations will seek out the benefits that tag containers bring.
8. Dashboards, dashboards, dashboards:  Dashboarding software is prime for a growth year (and some great options).  Advances from vendors such as Tableau should help analysts to focus on great visualizations of data and simplified reports.  2012 is the year for great dashboards.
9. Reporting multi-visit analytics: Most (if not all) web analytics platforms are insufficient in clickpath analysis in that they report on a single visitor on a single visit. For complex products such as home loans, network equipment, enterprise software, etc it usually takes multiple visits for a conversion to happen and sometimes requires multiple visitors.  In 2012 we hope to see better analysis tools to help analyze these complex and common paths.
10. Intranet analytics:  Web analytics for intranets have been an afterthought for web analysts for years.  Intranets are meant to be productivity tools and yet the data on their usage is typically inadequate.  Intranets have many dynamic functions such as document look ups, forms, phone directories, advanced searches.  More and more intranets are offering these complex functions to mobile users.  In 2012 we hope that web analysts will start to work with developers to track the complex functions of intranets and show how effective they are at improving employee productivity.
11. CMS integration:  Content management systems have come a long way with integrating web analytics in the last couple of years.  But there is still a far way to go.  At Extractable, we are hopeful that the leading CMS companies will continue to improve their integration with vendors such as Google Analytics, Webtrends, and Adobe.
12. Using visualization tools for better interpretation: The Visitor Flow report introduced by Google Analytics this year broadcast an interesting message to the analytics community.  Finding the right format for a report is critical to helping the team understand complex data sets and generate insights.  The more flexibility the analytics platforms give the analysts to report data the more effective the tools are.  In 2012, the we hope that the analytics platforms will continue to give new power to the analysts with powerful visualization tools.

One of the primary motivations for creating personas is building consensus internally within the design and product teams about customer goals and motivations. So why do personas go wrong so often? Why do people not use the personas that you labored to create from your research? Here are 5 tips to ensure the personas you create will indeed facilitate consensus.

1. Goldilocks Principle: Use not too much, not too little, just the right amount of detail to empathize with while avoiding conflict. Persona descriptions should be dense with details from user research. However, the details need some rounding out with characteristics and traits that will help make the persona seem like a real person that the design team could get to know. The trick is restraining the use of extraneous characteristics and traits that will either distract or create conflict within the design team. Stick to the facts and sprinkle in some stories and characteristics that you saw in your user research.
2. There is no “I” in team: Develop personas as a team. You need everyone’s buy-in for personas to be effective. The best way to build this buy-in is to involve everyone in the creation process of the personas. Gather the troops around a whiteboard and start mapping out the qualitative, quantitative, and colloquial knowledge about customers. This is a great time to also fill in gaps in team member’s knowledge about the customer research that has been done.
3. Use a 360-degree view: Think about the personas from different perspectives that the business has about the customer. Personas will be better accepted by stakeholders outside of the design and product teams if they resonate with how stakeholders view their customers. I’m not implying that personas should reflect back stakeholders’ current views of customers; rather, they should tell the story of the customer interacting with the stakeholder’s business unit. For example, the Support team will be more likely to use a persona if they see the persona description talks about how the persona struggled with the online user forum. Make the persona relevant to stakeholders.
4. Average isn’t good enough: Avoid filling personas in with average demographic content. While quantitative research gives you hard numbers about behavior, qualitative research helps you understand why that behavior happens. Don’t base personas solely off of the numbers. When creating personas, use dichotic analysis to map out the range of values and behaviors seen in research. Then make sure the personas represent that range. If your personas are just the market research with a little bit of character, people will feel as if there’s nothing new and interesting to digest about customers and will de-value the personas.
5. The power of one: Keep the focus of the personas narrow. Personas take a lot of work, and it can be tempting to try to use the same persona for different websites or products. However, trying to encompass all of the company’s products or offerings with one set of personas can dilute the effectiveness of the personas. Keep the persona narrow and tailored to the relevant design and product teams, and your team members will also stay focused on the website or product at hand.

It is essential for companies to start viewing their customer’s behavior from a holistic point of view – integrating data from both online and offline channels. However, many companies lack even a complete picture of what their customers are doing online. Companies need a way to measure how customers are engaging with social media and their website.

What do social media analytics track?

• Influence (Number of Twitter Followers/Facebook Fans/etc)
• Engagement (Number of Retweets/Likes/Comments/Interactions with Content/Number of Clicks)
• Social Media ROI (Conversions from a Social Media channel or Assisted by Social Media Channel)
• Audience Demographic and Location

Here’s a review of 3 social media analytics tools that integrate with Google Analytics – so there’s not excuse not to use them!

These tools range in functionality – so they can be used for one-man social marketing ninjas or larger PR teams. No website or brand is too small for social media analytics.

This feature analysis is based on the information found on the companies’ websites:

• http://sproutsocial.com/
• http://argylesocial.com/
• http://www.socialreport.com/

Find out more about Google Analytics integration on Google Analytics’ website: http://www.google.com/analytics/apps/results?category=Social%20Media%20Analytics

If the web analyst is an author, the dashboard is his/her novel.  It tells a story, with critical points in the story up to the interpretation of the readers.  And for the best analysts, each reader has seen the story through different perspectives.

Dashboards offer analysts a highly sophisticated tool set for telling a story which infers the intentions, the inclinations, the determination, the compulsion, and the reluctance of the visitors. I’m obviously over-romanticizing the dashboard, but this technology is a critical advancement in understanding how visitors interact with information and function.

While most analytic suites offer comprehensive reporting, dashboards are often created outside of the analytics tool for the following reasons:

1. Dashboards often require the integration of data from multiple sources (i.e. CRM, customer support systems, accounting software, etc)
2. Dashboards often have better tools for visualizing complex sets of data
3. Dashboards often have a better method of distribution of data (i.e. emails, intranets, portals, etc)

Finding the right dashboard platform, building the dashboards, and then implementing an analysis process which integrates the dashboards are significant undertakings (kind of like writing a novel).

Some of the toughest tasks are:

1. Exporting the data into a useful format
2. Combining the data from individual data sources into a common repository (usually a DB)
3. Associating fields from the various data sources (i.e. Customer ID with web site visitor)
4. Visualizing the data in a way that enables the viewer to make practical conclusions

It is an effort that requires both analytical skills and artistic senses.  While dashboards are typically great at showing what is (and isn’t) working on a website, it is difficult to gauge how successful the dashboards are at making an organization more insightful.

There are a lot of great options for dashboard platforms. Some of the tools we enjoy working with are listed here:

• Tableau
• Bime
• QlikView
• Trakkboard

Which dashboard platforms do you work with?

No one wakes up in the morning and says, “Today I’m going to navigate a website and click a button.” People have more important things to do in their lives, like helping their kids, or going on a hot date, or impressing their boss.

Sometimes in the pursuit of real world goals, people need to navigate a website and click a button, but we should never forget that they are doing so for some other reason – and before they make that click – that they must believe their real world goals will be aided.

People click a “buy now” button not because of how easy it was to navigate to find that button, not because the button was so cool looking, but because they believe something.

Sometimes these beliefs happen by reputation, sometimes by a feeling of affinity, sometimes by a feeling of validation, or any number of other impulses. And the way things look is a factor. But different people think and react in different ways.

If we limit ourselves to optimizing navigation, restructuring existing information, and reskinning interfaces – then we limit ourselves to being optimizers only. Just polishing stones. That’s not design, but merely decoration.

But if we dig deeper and define the “why” people take an action, instead of just allowing it, then we become the creators of reality. That’s design.

How do you make your reporting actionable?   We ponder this question often.

Good paid search firms have strong processes for reporting and acting on data. For example, they routinely look at phrases that have poor conversion rates and either alter the advertising message or eliminate the phrase.

But in web site optimization and web site redesigns, the processes and methodologies around making the data actionable are typically not as well defined.

Below are 7 simple tips that we have found help ensure actionable reporting:

1. Create Excel dashboards:  We first create site reports and dashboards in Excel which are much easier to change than web analytic reports or dashboard tools.  We review these reports with the team and ask each member if there is missing data that would help them improve the site or if we are reporting on data that they don’t know what to do with.  Eliminating data points is often as useful as adding data points.
2. Use Forecasts:  Predicting and justifying a lift in conversion is often the best way to get team members to start looking at creative ways to solve problems on the site.  When everyone is focused on solutions, they will start asking for different data points to act upon.
3. Use Segments:  On most sites, aggregate reporting doesn’t provide very useful information.  When we do reporting in the context of segments such as prospects, current leads, customers, loyal customers, investors, job seekers, employees, etc we find that the team members are often much more likely to identify critical trends that are actionable.
4. Encourage (or demand) question asking:  We sometimes have meetings where the objective is to educate the team on the behaviors of the visitors.  After data on site usage has been reported, we go around the room and have everyone ask questions.  This is a great technique for getting peoples creative juices flowing.  Often during the process, we uncover ways to improve the site.  No direction of questions is too ridiculous.
5. Use multiple data sources and use external research:  When a UX expert at a leading computer manufacturer read an article about how well the eye perceives the color green, he decided to run some MVT tests on their site by changing the Buy Now buttons to green.  The tests showed a favorable enough outcome that they rolled the new buttons out to the whole site.  This little improvement resulted in tens of millions of dollars in new sales and demonstrated that external data coupled with internal data can make a powerful catalyst.  Using multiple data sources helps show a more complete picture of visitor interactions.  Distribute data from systems such as CRM, eMail marketing, customer support, heatmaps, session playback, surveys, etc.
6. Analyze success and failure:  Often site reporting focuses on the number of visitors that actually convert.  This does not typically reveal the import insights that failure shows.  Analyzing both failure and success enables the team to identify methods for better serving visitors that don’t convert and helping them to convert.
7. Don’t just improve the sites:  The primary objective of site reporting is to improve each sites ability to serve visitors and to serve the organizations goals.   There should also be a routine process for developing actionable insights on how to improve the reporting that enables the team.  Dashboards, tracking, and data sources should be reviewed and improved quarterly.

We are always looking for ways to make site reporting more useful and more actionable.  We would love to hear from you on any tips you have for making data more actionable within your organization.  Send me an email at mryan@extractable.com.

As a data-driven firm, we are always looking for ways to understand exactly what people are doing on websites and what’s motivating them to look at or interact with certain parts of the website. Web analytics can give great visibility into that. However, click heatmaps give us better insight for large data sets with hundreds of interactions on any given page. Click heatmaps can tell a story that analytics or conventional usability testing never could.

*
What are click heatmaps?
Click heatmaps show the aggregate frequency of clicks on a page. Have you watched any Superman lately? Superman has “heat vision powers”…and would be a great UX professional. The orange/purple/red/yellow images that Superman sees are similar to the click heatmaps that we use to analyze websites. In a place on the page where more of the visitors have clicked, the click heatmap shows deep red. In a place on the page where fewer visitors have clicked, the click heatmap shows a light blue color.

Why do click heatmaps matter?
Click heatmaps are especially good at revealing what users are drawn to interact with on a specific page. This can help us understand two things:

• What content users are looking for or are most interested in
• What components of the visual design are helping to direct users to that content and what components of the visual design are distracting users from their goals

With that understanding, we can re-order content to make the most-used content more accessible, and we can tweak visual cues.

When should click heatmaps be used?

• With any page with heavy interaction, such as heavy searching, filtering, and sorting to understand how users are using tools and interpreting information
• When discovering opportunities in a new project by running click heatmaps on a current site
• When validating that a new design is directing people to the right content

Let’s recap on the benefits of click heatmaps:

• Cost-effective: More cost-effective than eye-tracking and has a 84-88% correlation in results (http://blog.clicktale.com/2009/11/23/eye-tracking-for-everyone/)
• Time-saving: Can be generated from the current site and does not require recruiting of users
• Session playback: Many click heatmap companies also include playbacks of sessions to discover the order and time in which interactions are performed by individual visitors
• Accessible: Easy for clients to understand the data
• Contextual: Looks at interactions in context of a whole page

*Photo credited to K2_UX at http://www.flickr.com/photos/k2_ux/5497508975/in/photostream/

Tweet

Posted in User Experience | No Comments »