# Trust and Dismissal [Image] This experience sits at the moment where people decide whether to accept outside guidance while doing focused work. Users are trying to write clearly without losing their own voice or momentum. For the business, this moment supports trust, long-term adoption, and the sense that the product earns a place in everyday work. 
We tested Grammarly’s suggestion experience inside Google Docs by showing users inline corrections and tone prompts in context. Participants were asked to imagine writing naturally and react to what the system suggested. The test used comprehension, desirability, success, sentiment, frequency, and intent to surface how clearly suggestions were understood, how confident people felt acting on them, and whether the experience fit into real writing habits. 
Experiences like this often break down quietly, not through errors but through hesitation. When people pause to question intent or accuracy, trust erodes even if the tool remains useful. Testing at this layer helps teams see where confidence fades, where authority feels unclear, and how small moments can shape long-term reliance on a product. * * * ## **User Needs & Business Goals** This experience tries to help people write more clearly while preserving their sense of authorship and intent. Users want to feel supported, not corrected, while the business aims to build trust and long-term adoption by staying helpful without feeling intrusive. **Audience** This concept was tested with adult knowledge workers in the United States who regularly write documents. Participants reviewed Grammarly’s suggestion experience inside Google Docs and were asked to imagine writing normally while reacting to inline suggestions, tone feedback, and corrections as they appeared. **User Needs** In this moment, users are focused on expressing ideas clearly while deciding how much outside guidance to accept. - The experience should feel easy to follow so users understand what a suggestion is asking them to consider **(intuitive).** - The experience should feel supportive without interrupting writing flow or attention **(usable).** - The experience should help users judge whether a suggestion fits their intent and voice **(insightful).** - The experience should feel trustworthy so users believe suggestions are made in their best interest **(credible).** - The experience should reinforce that the user stays in control of final decisions **(empowering).** Together, these needs balance clarity, trust, and control during a moment where confidence can easily slip. **Business Goals** From a business perspective, this experience supports long-term engagement and reliance on the product. - **Increase Feature Adoption** – Encourage consistent use of suggestions during everyday writing tasks. - **Build User Trust** – Strengthen confidence that recommendations are helpful and aligned with user intent. - **Improve Retention** – Keep users returning by making the experience feel supportive rather than corrective. - **Support Writing Efficiency** – Help users move faster without second-guessing decisions. - **Reinforce Product Value** – Demonstrate clear benefit in real work contexts, not just edge cases. These goals support durable value by aligning everyday usefulness with trust and continued use. * * * ## **Choose Metrics to Test Your Trust and Dismissal** This concept examines a writing support experience that appears while users are actively working. A design stack of UX metrics was selected by mapping core user needs to signals that show how people understand suggestions, decide whether to act on them, and incorporate the experience into real writing habits. The metrics used were comprehension, success, frequency, and intent. **Intuitive → Comprehension**
 People need to quickly understand what a suggestion is asking them to consider without breaking focus. Comprehension captures whether Grammarly’s prompts are clear in the moment they appear. This metric helps surface confusion or ambiguity that can undermine trust before a decision is made. **Empowering → Success**
 Users want to feel confident choosing whether to accept or dismiss a suggestion. Success reflects whether people feel capable using Grammarly to improve their writing without second-guessing themselves. This metric surfaces hesitation at the judgment moment, where confidence and control matter most. **Reliable → Frequency**
 For this experience to earn a place in daily work, it needs to feel dependable and worth returning to. Frequency reflects how often users expect to rely on Grammarly in this context. It signals whether the experience fits naturally into ongoing writing behavior. **Valuable → Intent**
 People ultimately decide where a tool belongs in their workflow. Intent captures whether users see Grammarly as something they would choose to use for real writing tasks. This metric connects perceived value to longer-term adoption decisions. * * * ## **Establish Hunches to Direct Your Testing** Teams usually have a sense of where an experience might be helping or getting in the way, but those assumptions are often fuzzy. Starting with hunches helps turn that uncertainty into clear questions that can be tested. Each hunch here reflects a risk worth examining before making decisions. **Example: Grammarly Google Docs**

Hunch

Question

UX Metric

Inline suggestions may be easy to notice, but users might not fully understand what Grammarly is asking them to change in the moment. If the intent behind a suggestion feels unclear, confidence can slip even if the correction seems minor.

How easy is it to understand what Grammarly is suggesting when it appears during writing?

Comprehension

Even when suggestions are clear, users may hesitate if they are unsure whether a change improves their writing or alters their voice. This hesitation could weaken confidence at the decision point.

How confident do you feel using Grammarly’s suggestions while writing?

Success

If Grammarly feels supportive rather than disruptive, users may expect to rely on it regularly during everyday writing tasks. If it feels distracting, usage may stay occasional.

How often would you expect to use Grammarly while writing in this context?

Frequency

Users may see Grammarly as useful in some situations but not others, depending on the type of writing they are doing. This could limit where they choose to adopt it.

Where would you be most likely to use Grammarly for your writing?

Intent

Taken together, these hunches aim to evaluate whether the experience supports understanding, confidence, and sustained use without undermining a user’s sense of control. * * * ## **Turn Hunches into Test Questions** Turning hunches into concrete questions helps reduce guesswork and keeps the test grounded in what people can actually respond to. Pairing each UX metric with a specific question type ensures the signals reflect real understanding, confidence, and intent rather than assumptions. **Comprehension (Likert scale)**
 *Question type:* Likert scale
 *Example:*
 How easy or difficult was it to understand what Grammarly was suggesting? **Success (Likert scale)**
 *Question type:* Likert scale
 *Example:*
 How confident do you feel using Grammarly’s suggestions while writing? **Frequency (Likert scale)**
 *Question type:* Likert scale
 *Example:*
 How often would you expect to use Grammarly while writing in Google Docs? **Intent (Multiple choice)**
 *Question type:* Multiple-choice
 *Example:*
 Where would you be most likely to use Grammarly for your writing? * * * ## **Calculate UX Metric Scores from User Feedback** This concept examined Grammarly’s suggestion experience inside Google Docs, where users decide whether to accept or dismiss guidance while actively writing. Participants were asked to imagine using Grammarly during real writing tasks and respond based on how the experience felt in that moment. The design stack included comprehension, success, frequency, and intent, combining attitudinal signals about confidence with indicators of expected behavior. - Very Good = 90% and above
 - Good = 70%–89%
 - Average = 50%–69%
 - Poor = 30%–49%
 - Very Poor = below 30% **Comprehension (72% — Good):**
 Most participants understood what Grammarly was suggesting when prompts appeared. The language and placement made sense quickly, allowing users to keep moving without stopping to decode intent. This indicates clarity at the surface level, even before trust is fully established. **Success (55% — Average):**
 Confidence dipped when users considered whether to act on a suggestion. While people could use the tool, some hesitated when deciding if a change truly improved their writing. This suggests the experience introduces doubt at the judgment moment rather than during interaction. **Frequency (87% — Good):**
 Participants expected to use Grammarly often in this context. The experience fits naturally into everyday writing habits and feels easy to keep around. Regular use does not appear to be a barrier. **Intent (77% — Good):**
 Users could clearly picture where Grammarly would fit into their workflow. They associated it with real writing tasks and practical scenarios, signaling perceived value even when confidence varies. [Image] Taken together, the scores point to an experience that earns presence and repeat use but struggles to consistently reinforce decision confidence. The dominant tension is not clarity or access, but trust at the moment of acceptance. This is a tool people keep open, even when they are not always sure they should follow it. Click here to check out the [raw survey data and UX metric scores for Grammarly’s GoogleDocs experience.](https://my.helio.app/report/01KD6X215VKAHS1S7DPB4154XY) * * * ## **Draw Signals from Your Design Stack** Here’s how signals were surfaced from Grammarly’s Google Docs test results by following five steps: **1\. Focus on poorly scoring or imbalanced metrics** [Image] The overall test score landed at 58% (Average). Frequency was the strongest metric, while Success lagged behind the rest. People can imagine using Grammarly often, but confidence drops when deciding whether to accept a suggestion. The weakest area is driven by hesitation at the judgment moment, not confusion about what the tool is doing.
Signal: Usage intent is strong, but trust softens when users must decide if a suggestion fits their intent. **2\. Identify patterns across metrics** [Image] Comprehension and Frequency reinforce each other. People understand the suggestions and see Grammarly as a regular companion while writing. That clarity does not fully carry into decision-making. Success and Desirability trail, revealing a tension between helpful presence and perceived authority. The friction sits in interpretation and confidence, not interaction. **3\. Determine if user needs are being met** [Image] - **Intuitive:** Met — Most users quickly grasp what Grammarly is suggesting. - **Usable:** Met — Suggestions fit into the writing flow without heavy disruption. - **Insightful:** Partially met — Some users struggle to tell why a suggestion is better for their specific intent. - **Credible:** Partially met — Trust varies depending on the type and tone of the suggestion. - **Empowering:** Partially met — Users feel supported, but not always confident owning the final choice. **4\. Compare outcomes to business goals** - **Increase Feature Adoption:** Supported — High frequency signals indicate regular use potential. - **Build User Trust:** At risk — Hesitation at acceptance moments weakens confidence. - **Improve Retention:** Partially supported — Ongoing use is likely, but long-term reliance depends on trust. - **Support Writing Efficiency:** Partially supported — Users move forward, but sometimes pause to second-guess. - **Reinforce Product Value:** Partially supported — Value is clear, but not consistently decisive. **5\. Surface signals & establish a direction** Signals derived from the data: [Image] - Users understand what Grammarly is doing, but not always why. - Regular use feels natural, even when confidence is uneven. - Trust breaks quietly at the moment of accepting or dismissing a suggestion. - The experience supports flow more than certainty. **Direction based on business context:**
 The evidence points toward an experience that succeeds by staying present rather than being forceful. The next gains likely come from reinforcing judgment and intent, not increasing visibility or volume of suggestions. This is a supportive writing experience that people keep around. Its main tension is not usefulness, but confidence at the moment of decision.