Every topic in the Buyer Criteria Analysis includes a win/loss record showing how often your brand comes out on top against competitors across the prompts within that topic. This gives you an aggregated view of performance so you can quickly see where you are consistently winning, where you are falling behind, and which topics are under the most competitive pressure in AI-generated answers.
What do wins, ties, and losses mean?
Each report runs queries on multiple AI models for every prompt within a topic. Gumshoe evaluates how your brand ranks across those models and compares that performance directly against each competitor. The brand that appears in the top position most consistently across models is considered the winner for that prompt.
- Win: your brand holds the top position most consistently across models for that prompt.
- Tie: your brand and a competitor appear equally often at the top across models, with no clear winner.
- Loss: a competitor holds that position most consistently instead.
Examples of wins, ties, and losses:
For example, your brand and two competitors are tracked in response to the prompt "What's the best project management tool for remote teams?" Across five AI models:
- Win: your brand ranks first in three models, and a competitor ranks first in the other two.
- Tie: both brands rank first in three models each, with no clear winner.
- Loss: a competitor ranks first in four models, and your brand ranks first in one.
How do I read the topic-level numbers?
The numbers shown in the topics list, such as 9 / 0 / 2, represent your total wins, ties, and losses across all prompts within that topic. These totals are recalculated each time a report run completes, so they always reflect the most current snapshot of your brand's performance.
Where does win/loss data appear?
Win/loss data is surfaced in several places within the Buyer Criteria Analysis:
- Topics list: scan performance across all topics at once, with wins, ties, and losses displayed side by side and color-coded for easy pattern recognition.
- Expanded topic view: drill into the individual prompts driving each outcome for a more granular breakdown.
- Radar view: win/loss data is combined with visibility scores so you can compare performance relative to competitors across all topics at once.
How do I read the expanded prompt cards?
When you expand a topic, each prompt is shown as a card that includes the persona asking the question, the question itself, and a visual representation of how brands ranked across models for that prompt. When there is a clear winner, the card displays a Top brand label identifying the brand that ranked first most consistently. Tied prompts omit that label.
By default, only a small number of prompt cards are shown to keep the view manageable. You can expand the full set to see every prompt contributing to the topic. Just click the "Show More" dropdown.
What is the competitive ranking section?
Below the prompt cards, each expanded topic includes a competitive ranking chart scoped to that topic specifically. It ranks all tracked brands by visibility percentage within the topic, which means results here may differ from your overall report performance.
The gap between you and the next competitor is one of the most useful signals in this section. A small gap usually means modest content or positioning improvements could shift the ranking. A large gap points to a more meaningful opportunity that may require deeper changes.
How do I use win/loss data to take action?
A good starting point is the topics where your losses meaningfully outweigh your wins, since those are the areas where competitors currently have a clear advantage in how AI models evaluate and recommend brands. From there, look at which competitors are winning those prompts and which personas are asking the underlying questions. That combination tends to point directly to the most important gaps to address.
Watch for a different pattern as well: a strong win record paired with relatively low visibility. In that case, your brand performs well when it is included in AI responses, but it is not being surfaced often enough. The right move is to expand the breadth of your content so your brand appears more consistently in relevant queries, rather than focusing on positioning.
For prompts where you are losing, the wording of the question itself is often one of the most valuable inputs available. The way those prompts are phrased reflects how AI models are interpreting user intent and evaluating brands. That language can be used directly to inform new content, refine existing pages, or adjust messaging to align more closely with how these systems are making decisions.
Click here to learn more about the Buyer Criteria Analysis.