What Makes a Good Prompt
Natural questions- Mirror how customers actually ask
- Use audience language
- Reflect real decision scenarios
- Broad enough for multiple brands
- Specific enough to stay relevant
- Consistently yields ~5–10 relevant results
- Matches your market/category
- Covers features, use cases, audiences
- Aligns with how customers discover/evaluate solutions
Prompt Examples by Industry
- SaaS/Software
- E-commerce/Retail
- Professional Services
- Consumer Brands
Project Management
- “What are the best project management tools for remote teams?”
- “Top project management software for agile development”
- “Best collaboration tools for distributed teams”
- “What are the leading CRM platforms for small businesses?”
- “Best sales automation tools for B2B companies”
- “Top customer management software for startups”
- “Best email marketing platforms for e-commerce”
- “Top social media management tools for agencies”
- “Leading marketing automation software for B2B”
Creating Your First Prompts
Step 1: Identify your categories- By use case: problems solved, tasks/goals, desired outcomes
- By audience: segments, industries/company sizes, evaluator roles
- By features/benefits: differentiators, most-valued features, benefits
1
Start broad
“What are the best [category] tools?”
2
Add specificity
“Best [category] for [audience]”
3
Focus on use cases
“Top [category] for [use case]”
4
Consider context
“Leading [category] for [context/constraint]”
- High-performing: mention you + relevant competitors; diverse, relevant answers; mirrors real decisions
- Low-performing: too broad/narrow; inconsistent/irrelevant; not customer-true
Managing Your Prompts
Add new prompts1
Navigate to Prompts
Open the Prompts section in your dashboard
2
Click 'Add new prompt'
Use the green button
3
Write your prompt
Enter a natural-language question
4
Save & activate
New prompts auto-enable and run in the next daily report
- Edit prompt text
- Enable/disable with toggles
- Monitor individual performance in Prompt Summary
- Delete underperformers
Changes appear in the next daily report. You cannot trigger reports immediately.
- Most accounts can add multiple prompts (limits vary by plan, but is up to 50)
- Contact mack@trakkr.ai to increase limits if needed
- Start with 5–10 high-quality prompts
- Remove consistent underperformers
- Focus on prompts that yield actionable competitive intelligence
Understanding Prompt Performance
Prompt Summary metrics- Visibility Score per Prompt: how well you rank when mentioned; shows strongest positioning; aligns prompts to brand strength
- Presence Score per Prompt: % of models that mention you; your share of voice; reveals awareness gaps
- Brand rankings: exact position per model; cross-model comparison; see which models favor you
- Competitor mentions: who models consider competitors; frequency; emerging threats/opportunities
- Response patterns: why ordering happens; context/framing that helps; language/positioning opportunities
Prompt Optimization Strategies
Improve low-performers If rarely mentioned:- Broaden prompt
- Change context (use cases/audiences)
- Check market relevance
- Consider brand awareness for that area
- Analyze top competitors’ positioning
- Refine targeting to strengths
- Adjust framing
- Highlight differentiation
- Audience: “Best [tool] for startups” vs “…for enterprises”
- Context: “Best [product] for beginners” vs “Professional [product] recommendations”
- Outcome: “Increase productivity” vs “Team collaboration”; “Cost reduction” vs “Growth”
- Industry trends: update for current focus; add emerging tech/methods; reflect seasonality
- Current events: new regulations/standards; industry changes; evolving needs
Common Prompt Issues
“My brand never appears”- Causes: too broad; weak training-data awareness; name variations; wrong audience/use case
- Fixes: get specific to your market; test known-fit prompts; check brand-name variants; assess online presence
- Why: different training data & cutoffs; specialization; brand/model biases
- Do: treat as normal; focus on trends; don’t over-optimize for one model
- Emerging/adjacent brands; related industries; international players—evaluate if truly competitive
Report Timing & Updates
When reports run- Daily automatic reports include all active prompts
- New prompts appear next day
- Edited prompts use the new text in subsequent reports
- No manual triggering (fixed daily schedule)
- Edit anytime in Prompts
- Changes take effect in the next daily report
- Historical data stays unchanged
- Trends update as new data arrives
Start broad to map the landscape, then layer specific prompts for segments and use cases.
Prompt Tags
Prompt tags let you group related prompts (e.g., by product line, service, campaign, or market) so you can measure performance at an aggregate level—not just per prompt. This is especially valuable when a single brand spans multiple products or services, because you can view how each area performs across all the questions you track. You create and manage tags on the Prompts page. Tags can be assigned to prompts one at a time when you’re editing individual rows, or applied in bulk to speed up large updates. Once tags are in place, the Tags section of your dashboard shows visibility aggregated by tag, giving you a clean, high-level read on how a product or theme is performing. You can also filter your Prompts table by tag to focus the view on a specific product, service, or initiative—making day-to-day analysis faster.1
Create tags on the Prompts page
Open the Prompts page, create tags you’ll use for grouping (e.g., “Enterprise”, “Payments”, “EU Market”). You can do this in each prompt row, or by using Manage Tags.
2
Assign tags to prompts
Apply tags manually (one at a time) while editing, or use bulk assignment to update many prompts at once.
3
Analyze by tag
Use the Tags section to see aggregate visibility and filter the Prompts table by tag for easier viewing.
Understanding Your Scores
Learn how prompt performance affects your overall visibility and presence scores
Reading Your Reports
Discover how to interpret your dashboard and track performance trends over time