The average Shopify store treats chat as a support channel. Someone asks a question, you answer it, the conversation ends. But every chat is also a sales conversation, and stores that understand this can see average order values climb without spending more on ads.
The difference isn't about being aggressive. It's about being helpful at the right moment with the right information. When a customer is already talking to you, they're telling you exactly what they need. Your job is to connect the dots.
Why Chat Is a High-Intent Channel
When someone opens chat, they're not browsing. They're stuck, curious, or one question away from buying. That makes chat higher intent than many other channels.
The problem is most stores answer the question and stop. Someone asks "Do you have this in blue?" and you say "Yes" instead of "Yes, and it's in stock. The matching belt is 20% off this week if you grab both." That's not pushy. That's useful.
Shopify tracks average order value across marketing channels in your analytics, and you can see how chat stacks up. If you're not actively working to increase AOV through chat, you're treating your highest-intent visitors the same as cold traffic.
Answer the Question, Then Show Them What Else Solves the Problem
The simplest way to increase AOV through chat without sounding like a used-car pitch is to answer what they asked, then add one relevant suggestion.
If someone asks whether a jacket is waterproof, the answer is yes or no. But the follow-up is where the sale can grow: "It is. If you're hiking in heavy rain, a lot of people pair it with our waterproof pants. Want me to pull those up for you?"
You're not selling. You're solving the same problem more completely. The customer already told you they care about staying dry. You're continuing that thread.
This works best when the suggestion is:
- A logical complement (the pants, the charger, the second bottle)
- Something that extends the use case (from "one trip" to "all season")
- A product that's frequently bought together (actual data, not a guess)
If you're using an AI agent connected to your Shopify catalog, it can recommend based on what other customers purchased alongside the item they're looking at. That means the suggestion is grounded in real behavior, not random cross-sell logic.
Use Bundles and Thresholds as Conversation Starters, Not Pop-Ups
Most stores throw a banner at the top of the site: "Free shipping over $75!" Then they never mention it again. In chat, that threshold becomes a conversation.
Hypothetical example: Someone adds a $60 item to their cart and opens chat to ask about sizing. If they're at $60 and your free shipping threshold is $75, that's a $15 gap. You can say: "That runs true to size. By the way, you're $15 away from free shipping. Want me to show you a couple of add-ons that put you over?"
It's specific, it's helpful, and it's tied to a concrete benefit. You're not inventing urgency. You're pointing out a deal they might miss.
The same logic applies to:
- Volume discounts ("Add one more and the per-unit price drops to...")
- Gift-with-purchase thresholds ("You're $10 away from the free travel case")
- Bundled pricing ("These three items are 15% off as a set")
The key is timing. Bring it up after you've answered their question, not instead of answering it. If you lead with the upsell, it sounds like a pitch. If you close with it, it sounds like advice.
Recommend Based on What They're Actually Trying to Do
Generic upsells fail because they ignore intent. "Customers also bought..." is noise unless it matches what the person is trying to accomplish.
Better: ask a clarifying question that reveals use case, then tailor the recommendation.
- "Are you buying this as a gift or for yourself?" (If gift: suggest gift wrap, a card, or a second item for the recipient.)
- "Is this your first time using this type of product?" (If yes: suggest the starter bundle or a how-to guide. If no: suggest the pro version or refill pack.)
- "How soon do you need it?" (If urgent: offer faster shipping. If flexible: suggest adding a backordered item they'd want next month.)
This is where AI support that's connected to your Shopify catalog can help. A human agent can ask these questions, but they're also juggling multiple other chats. An AI agent can ask, listen, query your product catalog in real time, and recommend the exact SKU that fits.
Kolton does this by pulling live product data, inventory, and variant details, so the recommendation isn't just relevant, it's actually in stock and ready to ship. That specificity is what makes the suggestion feel helpful instead of random.
Don't Discount Before They Ask
One of the fastest ways to lower AOV through chat is to offer a discount the moment someone hesitates. It trains customers to stall, and it costs you margin on sales that would have closed anyway.
Instead, use chat to remove friction. If someone is hovering on a product page and opens chat, they're not looking for 10% off. They're looking for a reason to trust the purchase. Answer that:
- "What questions do you have about this one?"
- "Anything I can clarify on size, material, or shipping?"
- "Want me to walk you through how this compares to the other option you were looking at?"
If they bring up price, that's different. But even then, the better move is often to reframe value (bundle, threshold, loyalty points, extended return window) rather than slashing the ticket.
Shopify's analytics let you track conversion rate and AOV by channel. If chat has a high conversion rate but a lower AOV than other channels, check whether you're discounting too early.
Handle Objections by Expanding the Cart, Not Shrinking the Price
When someone says "I'm not sure" or "Let me think about it," your next move determines whether AOV goes up or the sale dies.
The instinct is to offer a discount or back off. A better play is to ask what's holding them back, then solve it with a product or policy, not a price cut.
- "Not sure" often means they don't know if it'll work. Solution: point to reviews, sizing info, or your return policy. Or suggest adding the related accessory that makes it work better.
- "Let me think about it" often means they're comparing. Solution: ask what else they're considering and explain the difference. Or show them the bundle that includes both.
- "It's a bit more than I wanted to spend" might mean they're looking at the wrong SKU. Solution: show them the entry-level version and explain why other customers upgrade. Let them choose.
Each of those paths keeps the conversation going and gives you a chance to add value (and products) instead of subtracting margin.
For more on how to set boundaries around what your AI agent can offer or discount without approval, see Setting Guardrails for AI Support: What to Automate and What to Escalate.
Time Your Upsell to the Customer's Journey
Not every chat is the same moment in the journey. Someone asking a pre-purchase question is different from someone asking where their order is, and your upsell strategy should reflect that.
Pre-purchase: recommend and bundle
This is your highest-leverage moment. They haven't bought yet, so adding another item is frictionless. Use pre-purchase questions as signals:
- "Does this come with batteries?" → Suggest the battery pack or the bundle that includes them.
- "What's the difference between Model A and Model B?" → Explain, then ask which features matter most and recommend accordingly.
Post-purchase, pre-ship: add-on and upgrade
They've already bought, but the order hasn't shipped. This is the window for "Oh, and one more thing" purchases:
- "Your order ships tomorrow. Want to add anything else? No extra shipping cost."
- "A lot of people grab the cleaning kit with this. I can add it to your order if you want."
Post-delivery: replenish and expand
Once they've received and used the product, chat becomes the channel for refills, replacements, and related purchases:
- "How's the serum working for you? You're probably due for a refill soon. Want me to set that up?"
- "Since you have the starter set, some customers add the advanced kit next. I can show you what's in it."
Each of these stages has a different psychology. Pre-purchase is about solving the problem completely. Post-purchase is about convenience and avoiding a second shipping charge. Post-delivery is about continuation and trust.
Make It Easy to Say Yes in the Same Conversation
The reason many upsells fail in chat isn't the offer. It's the friction. If you recommend a product but the customer has to leave the chat, find it on the site, and add it manually, many won't bother.
Reduce that friction:
- Send a direct product link in the chat.
- If your chat platform supports it, show a product card with image, price, and an add-to-cart button.
- Offer to add it for them: "Want me to add that to your cart so you can check out together?"
Kolton sends product cards and links directly in the chat and can walk customers through adding items or adjusting their cart without leaving the conversation. That cuts the drop-off between "sounds good" and "actually purchased."
Track What's Working and Double Down
You can't improve what you don't measure. Shopify gives you average order value by channel in your marketing performance reports, but you need to go a layer deeper.
Track:
- AOV from chats where the agent recommended a second product vs. chats where they didn't
- Which product pairs or bundles get accepted most often
- Which suggestions get ignored or rejected
- Time of day, customer type (new vs. returning), and conversation length
If you see that customers who ask about shipping delays are more likely to add a second item when you suggest it, that's a pattern. Lean into it. If bundle offers during business hours convert but evening chats don't, adjust your approach by time.
Kolton's inbox tracks conversation history and outcomes, so you can see which recommendations your agent made and whether they converted. That feedback loop lets you refine the suggestions over time.
For more on how to manage this without doubling your support workload, see How to Scale a Shopify Store Without Hiring Support Staff.
The Real Difference Between Helpful and Pushy
Here's the test: if the customer says "no thanks" and you keep pushing, that's pushy. If you make one relevant suggestion, explain why it fits, and move on when they decline, that's helpful.
The goal isn't to upsell every chat. The goal is to make sure every customer knows what's available and why it might matter to them. Some will add it, some won't. The ones who do will spend more, and the ones who don't will still appreciate that you tried to help.
That's how you increase AOV through chat without sounding like you're reading from a script.
Where to Start
Pick one tactic from this list and test it for a week:
- Answer the question, then suggest one related product.
- Mention your free shipping or bundle threshold when someone is close.
- Ask a clarifying question about use case before recommending.
- Track AOV by chat vs. other channels and see where you stand.
You don't need to overhaul your entire chat strategy overnight. You need to stop treating chat like a question-and-answer service and start treating it like the high-intent sales channel it is.
If you want to see how an AI agent handles recommendations, upsells, and product questions, Kolton connects to your live Shopify catalog, answers in your brand voice, and works across web chat, Instagram, Facebook, email, and WhatsApp. See pricing and features at kolton.ai/pricing.
Key takeaways
- Chat is a high-intent channel. Every conversation is a chance to increase AOV by helping customers solve their problem more completely.
- Answer the question first, then add one relevant suggestion based on what the customer is actually trying to do.
- Use free shipping thresholds, bundles, and product pairs as conversation starters after you've answered their question, not instead of it.
- Don't offer discounts before they ask. Remove friction and add value by recommending products, not cutting price.
- Time your upsell to the customer's journey: recommend and bundle pre-purchase, add-on post-purchase, and replenish post-delivery.
- Track AOV by chat versus other channels and refine your recommendations based on what actually converts.
Frequently asked questions
- How do I suggest additional products without sounding like I'm just trying to upsell?
- Answer their question completely first, then add one relevant suggestion that solves the same problem or completes the use case. Frame it as helpful information, not a pitch. If they say no, move on. The key is relevance and timing, not repetition.
- What's the best time in a chat conversation to mention a bundle or free shipping threshold?
- After you've answered their main question and they're still engaged. If someone asks about sizing and you answer, that's the moment to say 'By the way, you're $12 from free shipping. Want me to show you a couple of add-ons?' Don't lead with it.
- Should I offer a discount if someone hesitates or says they need to think about it?
- Not right away. Ask what's holding them back first. Most hesitation is about fit, trust, or comparison, not price. Solve the real objection by pointing to reviews, explaining differences, or offering a better product match. Discounting too early trains customers to stall and lowers your margin.
- Can an AI agent actually increase AOV, or does this require a human?
- An AI agent connected to your live Shopify catalog can recommend products, suggest bundles, and mention thresholds in real time based on what the customer asks and what's in stock. Kolton does this across all channels and tracks what works. A human can do it too, but not at the same speed or volume.
- How do I track whether chat is actually increasing average order value?
- Shopify's marketing performance reports show AOV by channel. Compare AOV from chat to email, social, and organic traffic. Then go deeper: track which product recommendations convert, which bundles get accepted, and which conversations lead to bigger carts. Use that data to refine your suggestions.
Sources

About the author
Mario
Mario Horvat builds Kolton, the AI agent that helps Shopify stores turn conversations across chat, DMs and social comments into orders. He focuses on how the agent answers shoppers from a store's live catalog, stock and order data.
View profile →