Insights Analyst is a large language model that's trained on product reviews for your store. Once trained, Insights Analyst is ready to take any question you can think of about your customers' feedback.
We'll use this guide to give examples of what questions you can ask - and how you can use the knowledge you gain to inform your next questions.
💡 What you'll learn
What kinds of questions can Insights Analyst take? (jump to section)
Increasing your product knowledge (jump to section)
Content creation assistance (jump to section)
Improving customer support (jump to section)
💡 Prerequisites
A subscription to Insights Analyst will be required in order to use Insights Analyst.
What kinds of questions can Insights Analyst take?
Insights Analyst can take any question in natural language, but its responses will be focused on the product you select before interacting with the AI.
This targeted focus means you can be confident that the AI isn't pulling in learnings from irrelevant sources but it also means the best questions to ask are about the specific product or its reviews.
🔎 Here's an example of how an effective Insights Analyst conversation might go:
Ask Insights Analyst to summarize the negative feedback from each star rating.
Ask Insights Analyst to take the negative statements and reword them in ways that help alleviate customer concerns.
Next, ask Insights Analyst to create 5 variations of a customer support response to said concerns.
Refine the wording and content manually, or with further prompts to Insights Analyst.
Increasing your product knowledge
Insights Analyst is an encyclopedia of customer feedback on your products that you can speak with in natural language.
Some of the best questions to start with are the kinds of questions you want to know about your reviews, but would take a large amount of manual effort to answer.
Since Insights Analyst is always drawing from the same dataset for its knowledge, a best practice can even be to reframe the same question multiple times so that you can get different perspectives.
📕 Recommended questions to start with:
Can you summarize positive & negative reviews from each star rating?
Can you let me know what the most common positive and negative feedback from reviews are?
Can you show me what kind of feedback customers are most likely to leave a review for?
Can you quote some reviews from customers who were on the fence about an issue?
Content creation assistance
Insights Analyst is like a teammate and brand customer that offers to help you brainstorm and create content on demand.
The AI isn't just able to make content for a social media post, email & SMS marketing or even Octane AI quizzes but it can easily provide relevant quotes and SEO-optimized keywords provided directly from customer feedback.
Feel free to make your questions as complex as you'd like - Insights Analyst can handle it!
📕 Recommended questions to start with:
Can you create 5 drafts of a tweet about a sale on this product using the most popular positive keywords?
Can you write a blog post about this product that incorporates quotes, and addresses an issue customers have with a positive spin?
Can you show me a sales pitch that an in-store assistant would make about this product to a customer on the fence?
Can you show me keywords from each star rating that best represent why someone liked this product?
Improving customer support
Insights Analyst can empower your brand's customer support team to anticipate, iterate and refine responses based on real feedback.
This can be through high-level questions that helps your brand's customer support understand common customer feedback or through detailed questions that practice effective responses with the AI.
📕 Recommended questions to start with:
Can you quote an unhappy customer from a 1 star review and draft 5 sample support responses?
Can you draft instructions on how to use this product based on confusions that customers mentioned in their reviews?
Can you pull up 5 reviews that represent customers that were on the fence about the product, and also include responses for each one as if their reviews were support questions?
Is any of the negative customer feedback about issues that can be fixed with support, rather than product changes?