Online stores live and die by their support speed. A customer wondering "where is my order?" will not wait 48 hours for an email reply. They will open a dispute, leave a one-star review, or buy from a competitor.
According to Narvar's State of Returns report (2024), 73% of shoppers say the post-purchase experience determines whether they buy again. Support is not a cost center for e-commerce. It is a retention tool.
But hiring a support person at $3,500-$4,500 per month is premature if most of your questions are repetitive and answerable from documentation. Here is how to automate the five questions every shop gets before spending money on headcount.
The 5 Questions Every Online Store Gets
Scroll through any e-commerce support inbox. These five questions account for 60-70% of total volume.
1. "Where is my order?"
The most common e-commerce support question worldwide. Narvar (2024) reports that order tracking inquiries make up 35% of all post-purchase support tickets.
How to automate it: Write a knowledge base article titled "How to Track Your Order." Include step-by-step instructions: where to find the tracking number (order confirmation email), which carriers you use, and a link to the carrier's tracking page.
Add a section about delivery timelines. "Standard shipping: 3-5 business days. Express: 1-2 business days." Include a note about delays during peak periods.
When a customer asks the AI chatbot "where is my order?", it responds with the tracking instructions, links, and timeline expectations. No human needed.
2. "Can I return this?"
Return policies generate confusion because every store handles them differently. The questions cascade: What is the return window? Do I need the original packaging? Who pays for return shipping? How long until I get my refund?
How to automate it: Create a detailed return policy article. Cover every scenario:
- Return window (e.g., 30 days from delivery)
- Condition requirements (unworn, tags attached, original packaging)
- How to initiate a return (email, form, portal link)
- Return shipping cost (free, customer pays, deducted from refund)
- Refund timeline (5-10 business days after receiving the item)
- Exchange process
Make this the most thorough article in your knowledge base. An AI chatbot with a complete return policy article can handle 90% of return questions without human intervention.
3. "Do you ship to [country]?"
International shipping questions are binary. Either you ship there or you do not. But customers cannot always find the answer on your website.
How to automate it: Write a shipping information article. List every country you ship to. Include shipping costs per region and estimated delivery times. If you do not ship somewhere, say so explicitly. "We currently ship to the US, Canada, UK, and EU countries. We do not ship to Australia or Asia at this time."
The AI chatbot matches the customer's question to this article and gives a clear yes or no with the relevant details.
4. "What size should I get?"
Sizing questions are the top pre-purchase support topic for apparel and footwear stores. Getting the answer wrong means a return, which costs you $10-$20 in shipping and handling.
How to automate it: Create a sizing guide article with measurements in both inches and centimeters. Include a "between sizes" recommendation (e.g., "if you are between sizes, we recommend sizing up"). Add fit descriptions for each product category: slim fit, regular fit, relaxed fit.
If you have a sizing tool or quiz on your site, link to it directly in the article. The AI chatbot will include the link in its response.
5. "Is [product] in stock? When will it be back?"
Stock questions are tricky because inventory changes constantly. An AI chatbot cannot check your inventory in real time (unless you build a custom integration). But it can handle the general question.
How to automate it: Write an article about stock availability. Explain how to check stock on your website (look for the "Add to Cart" button; if it says "Sold Out," the item is unavailable). Include instructions for signing up for back-in-stock notifications if you offer that feature.
For the AI chatbot, set a fallback: "I cannot check real-time inventory, but you can see availability on the product page. If the item is sold out, sign up for a restock notification and we will email you when it is back."
Setting Up the Automation Stack
You need three things: a knowledge base, an AI chatbot, and a widget on your store.
Step 1: Write 10-15 articles (3-4 hours)
Start with the five questions above. Then add articles for:
- Payment methods accepted
- Discount codes and how to apply them
- Account creation and login issues
- Product care instructions
- Contact information and business hours
Each article should be 200-500 words. Short, specific, and easy to scan.
Step 2: Publish and connect AI (30 minutes)
Publish your articles in a knowledge base tool. Connect the AI chatbot. With Helpable, the AI reads your articles automatically. No training, no data upload, no configuration beyond the basics.
Test by asking each of the 5 common questions. Verify the AI gives accurate, helpful responses.
Step 3: Add the widget to your store (15 minutes)
Place the chat widget on every page of your store. Key placements:
- Product pages. Pre-purchase questions happen here.
- Cart and checkout. Discount code and payment questions happen here.
- Order confirmation page. Shipping and tracking questions start here.
- Contact page. Replace (or supplement) your email address with the widget.
A single JavaScript snippet handles all pages. Paste it once in your theme's header, and it appears everywhere.
The Time Savings Math
Assume your store gets 150 support messages per week. Each one takes an average of 8 minutes to read, research, and reply. That is 20 hours per week on support.
With a knowledge base and AI chatbot deflecting 50% of those messages:
- 75 messages handled by AI (zero human time)
- 75 messages still need a human (10 hours per week)
- 10 hours saved per week
At $20/hour (common for e-commerce support), that is $800/month in labor savings. The support tool costs $49/month. Net savings: $751/month.
At 60% deflection (achievable after 2-3 months of filling knowledge gaps): 90 messages automated, 60 need a human. That is 12 hours saved per week, or $960/month.
When Helpable Fits vs. When Tidio or Gorgias Fit
Not every e-commerce store needs the same tool. Here is an honest comparison.
Helpable fits when:
- You want a public knowledge base that ranks in Google and feeds your AI chatbot
- Your store is on any platform (Shopify, WooCommerce, Squarespace, custom)
- You care about flat-rate pricing without per-agent or per-conversation costs
- You want built-in NPS and CSAT surveys
- GDPR compliance and EU data residency matter to you
- You do not need deep Shopify order tracking inside the chat
Tidio fits when:
- Your store is on Shopify and you need native order tracking in the chat widget
- You want visual chatbot flows (if/then decision trees) for product recommendations
- You are willing to pay per AI conversation ($0.78 each with Lyro) for Shopify-specific automation
Gorgias fits when:
- You are on Shopify and process 1,000+ orders per day
- You need order management (refunds, cancellations, edits) directly inside the support tool
- You want deep Shopify, BigCommerce, or Magento integrations
- You have budget for $300+/month (Gorgias starts at $50/month for 300 tickets, scales to $750+ for larger volumes)
The honest summary: Helpable is the better choice for general e-commerce support with strong AI and a public knowledge base. Tidio and Gorgias are better if your primary need is Shopify-native order management inside the chat.
FAQ
Can an AI chatbot handle order-specific questions like "where is order #12345"?
Not without a custom integration. AI chatbots answer general questions from your knowledge base. For order-specific lookups, the chatbot should guide the customer to check their tracking link or connect them to a human agent.
How many articles do I need for an e-commerce knowledge base?
Start with 10-15 covering shipping, returns, payments, sizing, and common product questions. Most e-commerce stores reach full coverage with 25-40 articles. Add new ones whenever a question comes in that your AI cannot answer.
Will a chatbot annoy customers who want to talk to a human?
Only if there is no escalation option. Configure your chatbot to always offer a "talk to a human" button. The best setup: AI answers first, human available as a fallback. According to Salesforce (2024), 69% of customers prefer to use self-service for simple issues. The annoyance comes from being trapped in a bot loop with no escape, not from being offered help.
Is this approach only for big stores?
No. Stores doing 20+ orders per day already get enough support volume to benefit. If you process 5 orders per day, you might only get 3-5 support messages daily. At that volume, a knowledge base still helps (customers find answers via Google), but the AI chatbot's impact is smaller.
What about WhatsApp and Instagram support for e-commerce?
If your customers message you on WhatsApp or Instagram, you need a tool that supports those channels. Helpable currently focuses on website chat and does not offer WhatsApp or Instagram integration. Tidio and Crisp offer multichannel inbox options if social messaging is critical for your store.