Actually Asked

Product Introduction

1. Definition: Actually Asked is a self-improving, AI-powered FAQ (Frequently Asked Questions) widget and Shopify app. Technically, it is a SaaS (Software-as-a-Service) application that integrates via JavaScript snippet or native e-commerce blocks to provide dynamic, context-aware customer support directly on a website.
2. Core Value Proposition: It exists to automate and optimize customer self-service by answering visitor questions in real-time using a site's own content (pages, help center, product descriptions, policy documents) and autonomously expanding the FAQ section based on actual user queries. Its primary value is reducing repetitive support tickets, capturing purchase-intent questions, and creating a living knowledge base that evolves with customer needs.

Main Features

1. Inline, On-Brand FAQ Widget: The tool embeds as a native-looking section within a webpage, not a floating chat bubble. It uses CSS inheritance to automatically adopt the site's fonts, colors, and styling, ensuring a seamless user experience. Implementation involves pasting a two-line script (<div> container and <script> source) or using a dedicated block in platforms like Shopify.
2. Contextual AI Answering Engine: The system works by crawling and indexing a website's specified pages, help documentation, and uploaded files (PDF, Word). When a visitor asks a question in natural language or uses the "highlight-to-ask" feature, a retrieval-augmented generation (RAG) model queries this indexed knowledge base to provide accurate, sourced answers. It cites the specific page URL (e.g., /products/linen-shirt) for transparency.
3. Self-Learning & Autonomous FAQ Curation: The platform's core automation feature analyzes query logs to identify the most frequently asked questions. These high-demand questions are automatically drafted as new FAQ entries, which an admin can publish, dismiss, or edit. This creates a feedback loop where user behavior directly shapes and improves the support content.
4. Weekly Gap Analysis Report: A key analytical feature is the automated weekly email report. It provides metrics on total questions asked, answer success rate, and, crucially, lists questions the system could not answer based on existing content. This directly identifies gaps in a site's information architecture.
5. Multi-Platform Management (MCP & Integrations): Beyond a web dashboard, FAQ management is accessible via the Shopify admin panel and, uniquely, through AI assistants like Claude or ChatGPT using the Model Context Protocol (MCP). This allows for programmatic control. It offers native integrations for Shopify, WordPress (Custom HTML block), Webflow (Code Embed), Framer, Squarespace, and custom websites.

Problems Solved

1. Pain Point: Manual FAQ maintenance is reactive and stagnant. Businesses struggle to anticipate all customer questions, and static FAQs quickly become outdated, leading to increased support volume and abandoned carts.
2. Target Audience: The primary user personas are: E-commerce Store Owners/Managers (Shopify, WooCommerce), SaaS Product Marketers & Support Leads, Service-Based Business Owners (clinics, agencies, hospitality), and Web Developers/Designers implementing client solutions.
3. Use Cases: Essential for scenarios where visitors have pre-purchase or pre-booking questions: clarifying product specs (e.g., "Does this linen shrink?"), explaining service details (e.g., "What should I bring?"), confirming policy details (e.g., "Can I return after washing?"), and answering pricing/plan questions for SaaS.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros:
     • Zero-Config Styling: Automatically matches site design, requiring no CSS customization.
     • Autonomous Content Improvement: Truly "set-and-forget" capability that grows the FAQ organically.
     • Actionable Insights: The weekly gap report provides direct content strategy guidance.
     • Flexible Integration: Works across major website builders and headless setups via simple code.
     • Lead Capture: Can capture visitor emails for unanswered questions, turning support gaps into leads.
   • Cons:
     • Answer Limit on Free Tier: The free plan's 50 AI answers/month may be insufficient for moderately trafficked sites.
     • Dependence on Source Content: Answer quality is directly tied to the clarity and completeness of the indexed website content.
     • Limited Chat Interface: As an inline widget, it lacks the multi-turn, conversational depth of a dedicated AI chat agent.

2. Key Alternatives & Differentiation:

   • Traditional Chatbots (e.g., Intercom, Zendesk Answer Bot): These are typically separate chat interfaces focused on deflection to human agents. Actually Asked differentiates by being an inline FAQ that enhances the page content itself, not a separate conversation widget. It's generally more affordable and simpler to configure.
   • Static FAQ Plugins/Page Builders: Standard FAQ blocks are completely manual. Actually Asked is fundamentally different by being AI-driven and self-improving, automating both answer retrieval and content discovery.
   • Advanced AI Site Search (e.g., Algolia, Kapa): These are powerful search engines. Actually Asked is more focused and opinionated—it's not for general site search but specifically for Q&A extraction and FAQ management, with a stronger emphasis on autonomous content curation and styling cohesion.

Frequently Asked Questions (FAQ)

1. How does Actually Asked's AI ensure answers are accurate? Actually Asked uses a RAG (Retrieval-Augmented Generation) model that pulls answers directly from your website's own indexed content, help docs, and files. It cites the source page for every answer, ensuring responses are grounded in your verified information, not generating generic or potentially incorrect information.
2. Can I use Actually Asked on a non-Shopify website? Yes, Actually Asked is a platform-agnostic web app. You can install it on any website, including WordPress, Webflow, Squarespace, Framer, or a custom-coded site, by pasting a simple two-line JavaScript code snippet into your HTML.
3. What happens if the AI cannot answer a visitor's question? When the system cannot find a relevant answer in your indexed content, it will state it doesn't have an answer. Optionally, you can configure it to capture the visitor's email address for follow-up, turning the missed question into a support lead. These unanswered questions are prominently featured in the weekly gap report.
4. Is my website content and customer data kept private? According to its documentation, your site content is crawled to create a private index for generating answers. The service is designed to answer questions based on your public and provided documents, not to retain or expose personal customer data from queries.
5. How does the pricing work for high-traffic websites? Actually Asked uses a credit-based system where each AI-generated answer consumes a credit. Plans scale based on the number of monthly answer credits and the number of connected websites. High-traffic sites would need to subscribe to the Growth ($39/mo) or Premium ($99/mo) plans, which offer higher credit pools and multi-site support.