Ground techniques help teams bring real evidence into the conversation.
Once the problem is clear and outcomes are defined, the discussion needs something concrete to guide it. Without it, feedback relies on opinion, experience, or assumption. Signals connect the conversation to what is actually happening. They show how users behave, where they struggle, and what is or isn't working.
This step helps the group move from guessing to understanding.
Why this matters
When signals are missing, the review can start to loop. That usually shows up when:
-
Feedback becomes opinion-driven
-
People repeat the same points without progress
-
Assumptions are treated like facts
-
People defend their view instead of aligning
-
Decisions feel harder to support
When signals are present, the conversation changes. People react to something real, not just what they think. Signals do not need to be complex. They can be:
-
A moment where users hesitate
-
A pattern in behavior
-
A result from a quick test
-
A comment from user feedback
-
Something the team has already seen
What matters is that the signal reflects reality. The goal is not to prove everything. It is to give the group enough real evidence to move the conversation forward.
When you bring a signal in, say what kind it is. Each Design Signal is one of four types:
-
Need: what people need, from research and observation
-
Use: whether a design worked when people tried it
-
Prefer: which option people chose when they compared them
-
Adopt: whether people kept using it over time
The type tells the group what the signal can settle. A Use signal can show that a design works. It cannot show that people would pick it over another option. The evidence can come from a Helio test, your analytics, or the Glare MCP.
You are helping the conversation move from "I think" to "we're seeing."
When to use these techniques
Use Ground techniques when the conversation is relying too much on opinion. That usually happens when:
-
People repeat the same points without progress
-
Feedback is based on personal preference
-
Assumptions are treated like facts
-
Evidence is mentioned but not used
-
The group has no shared reference point
-
The conversation keeps circling
The goal is to give the group something real to respond to. Ground is working when signals start shaping the conversation. By the end of this step, the team should be able to say:
-
What signal matters
-
What the signal shows
-
How it relates to the outcome
-
Whether it changes the direction
-
What decision it helps support
This gives the review a shared reference point. Use the technique that matches what is happening in the room.
-
If opinions are repeating, bring in one clear signal.
-
If people cannot tell whether something improved, compare to a baseline.
-
If feedback sounds strong but unsupported, ask for the evidence behind it.
-
If a signal is mentioned but ignored, pause and let it shape the discussion.
Techniques
1. Bring in one clear signal
When a conversation is based only on opinion, it tends to repeat. People share views, respond to each other, and move in circles.
A single signal can change that.
It gives the group something real to focus on. It doesn't need to be perfect or complete. It just needs to reflect what users are actually doing. When that happens, the conversation gets more grounded and easier to move forward.
What to watch for
-
Repeated opinions without progress
-
People defending preferences
-
The same points coming up again
What this does
-
Anchors the conversation in reality
-
Reduces ambiguity
-
Helps the group converge
Example
"In testing, users hesitated at this step before continuing."
2. Compare to a baseline
It is hard to judge something without a reference point.
When people only see the current design, they react to it on its own. They may not know if it is better, worse, or just different. A baseline gives context. It shows what existed before and how things are changing. This helps the group see progress.
It also makes improvements easier to see, which builds confidence in the direction.
What to watch for
-
Feedback based on preference instead of progress
-
Trouble comparing ideas
-
Uncertainty about whether something is better
What this does
-
Gives context
-
Makes progress visible
-
Supports stronger decisions
Example
"This version improves on where users got stuck in the earlier flow."
3. Ask for the evidence behind feedback
Not all feedback comes from the same level of insight. Some comes from direct experience. Some comes from assumption. Unless someone asks, the two get treated the same.
Asking for evidence makes the difference clear. It encourages people to explain their thinking and connect it to something real. Over time, this improves the quality of the conversation.
What to watch for
-
Strong opinions without explanation
-
Conflicting feedback
-
Assumptions being treated as facts
What this does
-
Improves clarity
-
Strengthens feedback
-
Reduces noise
Example
"What makes you say that?"
"Have you seen that happen before?"
4. Pause when a signal shifts the conversation
Signals only help if the group has time to take them in.
In fast conversations, new information can come up and then get passed over. When that happens, it doesn't change the direction.
Pausing gives the group space to adjust. It lets people rethink what they've said and align around the new information.
What to watch for
-
Signals being mentioned but not discussed
-
The conversation moving too quickly
-
People continuing with earlier assumptions
What this does
-
Helps signals land
-
Improves alignment
-
Supports better decisions
Example
"Given that, does this still feel like the right direction?"
Bring real signals into the conversation. When the discussion is grounded in what is actually happening, feedback gets clearer, alignment comes faster, and decisions are easier to make.
Signals don't need to be perfect. They just need to be real enough to move the conversation forward.