When an assistant mentions your SaaS, the words around your name shape the buyer's impression. “Easy to set up but expensive for larger lists” leads to a very different decision than “the most affordable option for growing teams”. Sentiment and perception analysis in Narrashift captures these descriptions systematically, so you know not only whether AI talks about you, but what it says.
Strengths and weaknesses with evidence
For every analyzed answer, Narrashift identifies explicit evaluations of your product and your competitors: an attribute such as pricing, ease of use, deliverability or support, together with a contiguous verbatim quote from the answer. If the analysis cannot quote a statement exactly, it is not recorded. This strict rule prevents paraphrased or invented claims from creeping into your reports.
Attributes are aggregated across answers, so you see which strengths are mentioned most often and which weaknesses keep recurring. Each item links to the answers it came from. The overview lists the top points under “What stands out” and “Room to improve”, and the brand comparison section shows your pros and cons next to your key competitors.
Sentiment you can trust
Each mention of a brand is classified as positive, negative, neutral or mixed, and answers that do not mention you are counted separately rather than as neutral. The sentiment chart shows the distribution for the selected period and provider, so a shift — for example, more mixed descriptions after a pricing change — is easy to notice.
Because sentiment is stored per answer and per provider, you can see whether a negative description comes from one model or from all of them. A weakness repeated by every assistant usually points to a widely published source; a weakness in a single model may come from outdated training data.
From perception to action
Perception data feeds directly into optimization. Opportunity finder groups recurring weaknesses into negative perception opportunities, each with the quoted answers as evidence. The suggested action is never to argue with the AI: it is to check whether the statement is accurate and current, update or publish the relevant information with a date if it is outdated, and document the limitation honestly if it is true.
Positive attributes are just as valuable. If assistants consistently praise your onboarding, that is a message to reinforce on your website and in your documentation, and a strength to highlight in comparison content.
Watching the narrative change
After you update public information, the Impact tracker compares sentiment for the affected prompts before and after the change. The goal of the whole product is in the name: not only to track the narrative about your SaaS, but to help you shift it — with evidence at every step.
Perception across providers and competitors
Perception is relative. A weakness matters more when competitors are praised for the opposite, and a strength matters less if every product in your category gets the same compliment. That is why Narrashift extracts strengths and weaknesses for competitors too, using the same rules and the same quoting requirement. In the brand comparison you can see whether assistants describe a competitor as easier, cheaper or better supported, and which answers say so.
Branded and unbranded questions are analyzed separately as well. On branded questions buyers already know your name and want details, so the descriptions there reveal how assistants summarize your product. On unbranded questions the description decides whether you make the shortlist at all. Comparing the two often shows that a product is described well when asked about directly, but loses on the generic questions where most buyers start.