Latest / 5 Minute UX / Selecting Tools for Design Testing
Transcript
- 0:08This lesson is Selecting Tools for Design Testing. Let's get
- 0:12started.
- 0:19When you need to select the appropriate design testing tool,
- 0:22start by identifying the three core questions of the tool
- 0:24selection decision framework. The first question asks what you
- 0:27are essentially trying to discover, which determines
- 0:29whether you need visual behavior data or information architecture
- 0:32insights. The second question focuses on what forms of data
- 0:35you want to capture, distinguishing between cursor
- 0:37movements and heat maps versus longer recordings of user
- 0:39feedback. The third question considers who will view the
- 0:42results, ensuring the data type is most effective for both team
- 0:45members and executive sponsors. This framework exists because
- 0:47the choice is not arbitrary; it depends on whether the objective
- 0:50is visualizing behavior, testing information architecture, or
- 0:53conducting structured usability tasks. For example, if the goal
- 0:56is a visual representation of behaviors discoverable via web
- 0:58analytics, you use a tool like ClickTale on live versions of
- 1:01the site. Conversely, if the goal is to build out the
- 1:04information architecture, you use OptimalSort to create a
- 1:06remote card sorting test. Determining what is essential
- 1:09allows you to use that information to choose the best
- 1:11tool for the specific research objective. When in doubt,
- 1:13choose the tool that gets the best answers to the project's
- 1:16"burning questions." This approach ensures that the
- 1:18selected tool aligns with the specific research goals, data
- 1:21types, and audience needs of the project. Once these three
- 1:24questions are answered, you can match the goal to the
- 1:26appropriate tool category.
- 1:33Once you have identified your core question, you map that goal
- 1:36to a specific tool category. Behavior visualization tools
- 1:39like ClickTale, Crazy Egg, or UserFly turn raw web analytics
- 1:42into heat maps and click path videos. You use these on live
- 1:44sites to spot design issues or run A/B tests on two different
- 1:47versions. They show you exactly where users are clicking and
- 1:50where they are hesitating. Usability testing platforms like
- 1:53Chalkmark, Loop11, UserZoom, UserTesting.com, or Usabilla
- 1:56serve a different purpose. They let you set up multi-step tasks
- 1:59and related questions for a larger sample of users. The
- 2:02output is data on user success rates and specific behaviors for
- 2:05those tasks. This is where you get the quantitative evidence
- 2:08that a particular workflow works or fails. Information
- 2:11architecture tools like OptimalSort or Treejack focus on
- 2:14navigational and category structure. You use these when
- 2:16page designs are not yet available or not the primary
- 2:19concern. They are built for card sorting and testing how users
- 2:22intuitively group content. If your goal is to build out the
- 2:25information architecture before the visual design is finalized,
- 2:28this is the category you choose. Each category answers a distinct
- 2:31type of question. Selecting the right one ensures your data
- 2:34directly addresses the project's burning questions rather than
- 2:37providing generic metrics.
- 2:44Here’s the framework applied in practice. Imagine your team is
- 2:47debating whether to launch a new navigation menu. The first
- 2:50question is what you are essentially trying to discover.
- 2:52If the goal is a visual representation of behaviors
- 2:54discoverable via web analytics, you don't need a full usability
- 2:57study. You select a tool like ClickTale or Crazy Egg to run on
- 3:00the live site. This gives you heat maps of click paths
- 3:02immediately, showing exactly where users hesitate or abandon
- 3:04the flow. It’s the fastest way to see if the current design is
- 3:07actually working before you change a single pixel. Now,
- 3:10suppose the problem isn't the visual design, but the structure
- 3:12itself. You’re building out the information architecture before
- 3:15any page designs exist. In that case, a heat map is useless
- 3:17because there’s no layout to analyze. Instead, you use
- 3:20OptimalSort to create a remote card sorting test. Participants
- 3:22arrange categories in their own mental model, revealing the
- 3:25logical structure that supports your navigation. This step is
- 3:28critical because it tests the skeleton of the site, not the
- 3:30skin. Finally, consider a scenario where the structure is
- 3:33solid, but you need to verify that users can actually complete
- 3:35a complex task. You want to test multi-step tasks with a larger
- 3:38sample for success rates. This is where you turn to platforms
- 3:41like UserTesting.com or Loop11. These tools set up structured
- 3:43usability tests with related questions, capturing not just
- 3:46clicks, but the reasoning behind them. You get a larger sample of
- 3:49data showing user success rates and behavior for specific
- 3:51questions and tasks. This depth of evidence is what moves a
- 3:54design decision from a guess to a validated choice. The key is
- 3:56to determine what is essential for your specific burning
- 3:59questions and let that drive the tool selection. You are applying
- 4:01the framework to choose between ClickTale, OptimalSort, or
- 4:04UserTesting.com based on a specific research objective.
- 4:06This ensures your data matches your needs. When you get the
- 4:09tool wrong, the data won't answer the question you actually
- 4:11have.
- 4:17Most practitioners reach for the tool they’ve used before, not
- 4:20the one that answers the project’s burning questions.
- 4:23That familiarity bias bypasses the core decision framework and
- 4:26leaves the research misaligned with the actual objective. If
- 4:28the goal is visualizing behavior on a live site, ClickTale is the
- 4:31right choice, regardless of whether you’ve used
- 4:34UserTesting.com in the past. The second mistake is ignoring
- 4:36who will view the results. You need to select data formats that
- 4:39effectively communicate to executive sponsors, not just the
- 4:42team. A heat map might be useful for engineers, but it often
- 4:45fails to convey the business impact that stakeholders need to
- 4:48approve the next design phase. Finally, overlooking logistical
- 4:51constraints delays the entire process. Selecting a tool
- 4:53requires working with internal IT teams to ensure appropriate
- 4:56security, permissions, and processes for adding team
- 4:58members to licenses. If you skip this step, the tool sits unused
- 5:01while the project timeline slips. Choose the tool that
- 5:04gets the best answers to the project’s burning questions, set
- 5:07the right expectations with the team, and confirm IT permissions
- 5:10before you start. When you next face a design decision, ask what
- 5:13you are essentially trying to discover, and let that answer
- 5:16drive your tool selection.
- 5:20Welcome to The Take. I'm joined by a fellow practitioner and
- 5:23together we're going to dig into Selecting Tools for Design
- 5:26Testing — what it means in practice and what's worth
- 5:28remembering.
- 5:30[HOST] The part about choosing tools for "burning questions"
- 5:33hit hard because I still catch myself defaulting to ClickTale.
- 5:36I just assume it's the safest bet for any visual issue.
- 5:39[LEARNER] That familiarity bias is a trap. I used to do the
- 5:42same, but I realized a heat map tells me where they clicked, not
- 5:46why. [HOST] Exactly. If the structure is broken, a heat map
- 5:49is just noise. You need OptimalSort to test the skeleton
- 5:52before you worry about the skin. [LEARNER] True, but getting
- 5:55executives to care about card sorting is a different beast.
- 5:58They wanna see the finished product, not the abstract
- 6:01categories. [HOST] That’s the audience mismatch. Heat maps
- 6:04feel tangible to them, but they often miss the business impact
- 6:07of a poor structure. [LEARNER] So how do you bridge that gap? I
- 6:11usually have to translate the sorting data into potential
- 6:14revenue loss to get their attention. [HOST] You have to
- 6:17frame it as risk mitigation. Show them that a bad structure
- 6:20wastes their marketing budget on traffic that can’t convert.
- 6:23[LEARNER] That works. But what about the logistical side? IT
- 6:26departments can make a simple tool selection a month-long
- 6:29project. [HOST] Totally. If you skip the permissions check, the
- 6:32tool sits idle. I’ve seen projects slip because we forgot
- 6:35to request API access early. [LEARNER] It’s frustrating. You
- 6:39pick the perfect tool, then spend three weeks waiting for a
- 6:42single license key from the security team. [HOST] It’s a
- 6:45common pain point. You have to treat the tool selection as a
- 6:48procurement process, not just a design decision. [LEARNER]
- 6:51Agreed. I’m starting to use AI to draft the security
- 6:54justification documents now. It speeds up the approval phase
- 6:57significantly. [HOST] That’s a smart workaround. AI handles the
- 7:00repetitive compliance text, so you can focus on the actual
- 7:03research strategy. [LEARNER] Yes. It takes the friction out
- 7:06of the bureaucratic layer, but it doesn't solve the core
- 7:09problem of choosing the right tool. [HOST] Right. The AI helps
- 7:13with the paperwork, but your judgment still has to drive the
- 7:16selection. Don't let the process dictate the research. [LEARNER]
- 7:19Good reminder. I almost got lost in the vendor comparison matrix
- 7:23last week. [HOST] Keep the burning question at the front of
- 7:26your mind. The tool is just a means to an answer, not the goal
- 7:29itself. [LEARNER] I'll take that to heart. It’s easy to forget
- 7:32that the data serves the decision, not the other way
- 7:35around. [HOST] Precisely. If the data doesn't answer the specific
- 7:39question, it’s just expensive noise. Always start with the
- 7:42objective.