TL;DR:

• Shoppers won’t wait for help—so they leave. Without instant answers, you’re losing sales to hesitation and unanswered questions.
• AI shopping assistants step in right away. They guide, recommend, and resolve concerns in real time, 24/7.
• They boost conversions and AOV. Brands like Pepper and TUSHY saw up to 62% more conversions after adding one.
• They go beyond chatbots. AI assistants proactively sell, using browsing behavior to tailor messages and close more carts.
• Gorgias Shopping Assistant is built for Shopify brands. It starts conversations, mirrors your brand voice, uses browsing data, suggests products, and sends dynamic discounts.

Shoppers aren’t always going to reach out and ask the questions they have, especially if they’re going to have to wait for a response from a CX team. 

That means you’re losing sales to friction, indecision, or information gaps. 

In 2025, the average cart abandonment rate is 70.19%. But if you can find an AI tool that doubles as a support and sales agent, it could make all the difference. 

Gorgias’s Shopping Assistant, for example, has brought a 62% uplift in conversion rate for brands that implement it.

Ahead, learn where you can leverage an AI shopping assistant to increase conversions and craft better purchase experiences. 

What is an AI shopping assistant? 

An AI shopping assistant is a chat tool powered by AI to provide pre-sales support for shoppers. It can answer questions, make product recommendations, and help guide shoppers in the right direction if they’re stuck. 

Gorgias's Shopping Assistant is a powerful, hyper-personalized AI tool built for Shopify brands. Unlike other AI tools, Shopping Assistant starts conversations with customers, not the other way around. It’s uniquely tailored for each customer by tracking browsing behavior during each session and remembering what shoppers say, keeping conversations natural and recommendations relevant. 

It’ll also chat with shoppers in your own brand voice, as its responses are pulled right from the knowledge you feed it. 

At which point do brands lose sales in the customer journey?

The stages of the customer journey where common drop-off points occur for brands that lack proactive support include: 

• Discovery (casual browsing)
• Interested (considering making a purchase)
• Ready to buy (strong purchase intent)  

1) Discovery (casual browsing) 

There’s a big chance that shoppers—especially first-timers—have questions, but aren’t willing to wait for a human to get back to them. And when your CX team is off the clock? Customers will likely leave altogether.

An AI shopping assistant can help you engage customers right away, even outside your business hours.

Bra brand Pepper uses Gorgias Shopping Assistant to help shoppers find their perfect size. When it detects hesitation, Shopping Assistant points customers to the sizing guide.

This proactive approach creates an easy path for conversation and sets the precedent that any questions will be answered immediately, providing a better––and less confusing––experience. 

“With Shopping Assistant, we’re not just putting information in our customers’ hands; we’re putting bras in their hands,” says Gabrielle McWhirter, CX Operations Lead at Pepper. 

Impact these metrics 📈

For shoppers in the Discovery stage, using a Shopping Assistant boosts clicks and time on site and reduces bounce rate. It does this by surfacing specific questions on relevant product pages. Pepper boosted their conversion rate by 19% with Gorgias Shopping Assistant.

Read more: How Pepper’s AI Agent automates 54% of support and converts 19% of conversations

2) Interested (considering making a purchase)  

In a retail environment, a salesperson can give shoppers recommendations by asking a few questions, especially if they’re unsure of what to buy. 

AI shopping assistants have the ability to mirror those in-person shopping experiences by interacting with customers in real-time to help them find their perfect match.

Shoppers can give as much (e.g., “Help! What dress is suitable for a wedding reception?”) or as little information as they’d like, and the AI shopping assistant will do the rest. 

It’s possible even for questions that are slightly vague, like a customer who types in “how to make up” without any other context:

For example, jewelry shop Caitlyn Minimalist uses Shopping Assistant to recommend products, engaging interested customers and bringing them closer to a purchase.

“As a result of Shopping Assistant, we've seen a measurable lift in AOV through more meaningful customer interactions,” says Anthony Ponce, Head of Customer Experience at Caitlyn Minimalist. 

“Our clients are provided the right information at the right time, creating a seamless experience that builds trust and drives confident purchases." 

Impact these metrics 📈

According to data from Gorgias, email is the highest volume support channel, with ~25% of that tied to pre-sales. AI shopping assistants tackle these pre-sales asks and also upsell by recommending complementary products. This can lead to a boost in average order value (AOV) and conversion rate. 

Read more: How Caitlyn Minimalist uses Shopping Assistant to turn single purchases into jewelry collections

3) Ready to buy (strong purchase intent) 

The main reasons customers abandoned a cart in 2025 include:

• Extra costs (like shipping, fees, taxes) - 39% 
• Delivery times were too slow - 21% 
• Checkout was too complicated or long - 18% 
• Didn’t like return policy - 15% 

An AI shopping assistant can mitigate or resolve these issues. They resolve crucial questions—like delivery time or return policies—that need in-the-moment answers. By alleviating pre-sale concerns, they give customers the confidence to make a purchase.

For example, bidet brand TUSHY leverages Shopping Assistant to answer questions about toilet compatibility that might flush a pending sale. 

Aside from quelling customer concerns, Shopping Assistant can also send discount codes to close deals. Unlike general discount codes you find across the internet, these discounts are uniquely generated for each customer, keeping them engaged and on your site.

Impact these metrics 📈

AI shopping assistants can reduce cart abandonment rate and increase conversion rate. Gorgias Shopping Assistant adjusts to your sales strategy by sending customers discount codes that can be the final nudge to checkout.

Read more: How TUSHY uses Shopping Assistant to drive 190% higher chat conversion rate by building customer confidence

Why traditional chatbots don’t cut it anymore

Most AI tools are built just for support. They deflect tickets and answer FAQs, but they’re not built to sell. 

Shopping Assistant proves that support teams can also drive revenue by upselling, suggesting exchanges, and giving shoppers the confidence to try a brand for the first time (or to give it another shot).

Gorgias’s AI Shopping Assistant uses context-based decision making and looks for specific behavioral signals: 

• Products viewed - Which product pages have they visited?
• Pages views - What pages have they explored, like FAQs, reviews or categories?
• Current page - Which page are they currently browsing?
• Purchase history – What have they bought from this brand before?
• Cart data - What items are currently in their cart?

Improve conversions with Gorgias Shopping Assistant 

Ultimately, the cost of not adopting AI can be higher than the investment of implementing it. 77.2% of ecommerce professionals use AI to improve their work. Why not extend those benefits to your customers? 

AI Shopping Assistants help you create better customer experiences overall. These tools help reduce customer effort, increase average order value, save would-be-lost sales, and create more customer touchpoints.  

Hire the always-on Shopping Assistant that never misses a sale.

{{lead-magnet-1}}

TL;DR:

• Shopping Assistant is your new AI sales closer. It jumps in when shoppers hesitate, delivering real-time, personalized support that drives conversions.
• It boosts revenue by acting like your best salesperson. It knows when to recommend, upsell, or offer a discount, without being pushy.
• It remembers what shoppers do mid-session. That means smarter conversations, better suggestions, and bigger order values.
• It’s already delivering results. Brands using Shopping Assistant are seeing 62% more conversions, 10% higher AOV, and 5x ROI.
• You control the strategy. Customize how it promotes discounts, when it steps in, and how it speaks.

What’s the common factor between shoppers debating between products and considering a splurge? Hesitation. 

Today’s shoppers are overwhelmed with choices. They don’t want to be left to figure things out on their own. They want guidance.

But most brands are missing that crucial piece of the puzzle. They lack a strategy that accompanies shoppers on their journey. A tool that encourages shoppers to proceed to checkout. And, ultimately, a customer experience devoid of a sales approach.

That’s why we built Shopping Assistant, an AI Agent that proactively engages browsers, offers context-aware product recommendations, and turns hesitation into conversions in real time.

And it’s working. Brands using Shopping Assistant are seeing a 62% uplift in conversion, 10% higher average order value, and 5x ROI.

Here’s a closer look at what’s behind the magic.

AI-powered shopping, built for sales

Most traditional chatbots passively wait for questions and deliver answers that aren’t personalized to each shopper's preferences. 

Unlike these bots, Shopping Assistant reads real-time signals like pages viewed, cart contents, and conversation tone. This results in a solution that not only offers support but also offers personalized, proactive selling. This enables Shopping Assistant to continuously refine and adjust its playbook, evolving with each shopper as their journey matures.

Here’s how Shopping Assistant engages with customers across the shopping journey:

1. Discovery: Gathers preferences and educates.
2. Interested: Makes targeted product suggestions.
3. Ready to Buy: Assists with checkout, nudges purchases with discounts.

Take this example below. When a customer vaguely asks “how to make up,” Shopping Assistant interprets it as a sign of interest in makeup products and recommends a starter kit.

Personalization that understands you

Where traditional bots reset with every message, Shopping Assistant does the opposite. It has built-in context-aware intelligence that remembers what shoppers have clicked, viewed, and added to their cart during a session. 

This enables natural, relevant, and persuasive conversations that truly resonate with each shopper. It goes beyond reading messages and observes behavior to adapt its responses.

That means it knows if someone has:

• Viewed three red lipsticks but didn’t add to cart
• Repeatedly checked sizing info for the same dress
• Added two serums to their cart after browsing dry skin FAQs

With plenty of context to work with, Shopping Assistant is not only smarter but also more profitable than the average chatbot. It drives more conversions with product recommendations and lifts average order value with timely upsells based on what’s been added to the cart or viewed. 

Here’s what it looks like in action: When a customer engages through a product page, Shopping Assistant recommends a matching outfit, suggesting it’s aware of alternate product variants and the customer's likely interest in that style.

Dynamic discounts that convert without undercutting

Promotions are powerful, but they’re not one-size-fits-all. 

With Shopping Assistant, merchants can define their discount strategy to align with their brand. These strategies can range from offering no deals to using aggressive promotions. 

Once the strategy is set, Shopping Assistant waits for hesitation and customer intent to trigger a discount, firing it at the most conversion-worthy moment.

Proactive engagement in 3 ways

Shopping Assistant initiates conversations. It’s built to engage shoppers, spotting when they linger or show signs of confusion, stepping in with timely, personalized help.

Every second counts in ecommerce. If a shopper pauses on a product page or is left scrolling through an endless search results page, Shopping Assistant detects it in real-time and reaches out with a relevant prompt like:

• “Need help picking the right shade?”
• “Want to know our return policy before you buy?”

Here’s how Shopping Assistant reduces drop-off, builds confidence, and drives faster decision-making in three different ways.

1. Suggested product questions 

Shopping Assistant automatically triggers commonly asked questions depending on the product currently being viewed. In one click, shoppers can get the answer to the question they’re curious about. This combats hesitation caused by a lack of information, resulting in more confident conversions.

2. Ask Anything Input

When shoppers land on the homepage, it’s easy to become overwhelmed and not know where to navigate. The Ask Anything Input provides an easy way to start a conversation with Shopping Assistant and get the guidance they need.

Shopping Assistant can refine its response to the customer based on the page context. For example, when the customer is on a product page, Shopping Assistant knows exactly what product is being asked about.

3. Trigger on Search

Shopping Assistant can step in to offer pinpointed help based on a shopper’s search query. Instead of scrolling through a results page, Shopping Assistant triggers a message based on what the shopper entered, offering an easier and faster way to find what they need.

Smart recommendations and relevant upsells

Shopping Assistant’s suggestions are rooted in real context: what the shopper has viewed, added to cart, or asked about. Whether they’re exploring a specific product line or revisiting a category they’ve shown interest in, Shopping Assistant delivers relevant upsells and complementary items that make sense for the customer.

This personalized approach to upselling increases cart size without feeling forced—it’s smart, seamless, and sales-driven.

Shopping Assistant can even turn vague product questions into upsell opportunities. By asking questions, it learns more about an individual to come up with recommendations that best fit their preferences.

Try Shopping Assistant today

Shopping Assistant is transforming the way shoppers engage and helping ecommerce brands sell more effectively. Through smarter conversations and real-time personalization, it turns every interaction into an opportunity to convert, build trust, and drive revenue.

{{lead-magnet-1}}

TL;DR:

• Don't just turn it on, coach it. Treating your AI Agent like a team member, not a plug-and-play tool, is what makes it truly helpful and memorable.
• One owner and one hour a week is enough to make a big impact. SuitShop’s Katy Eriks runs QA and training solo, using a repeatable system to log feedback and improve performance weekly.
• Don't forget to pause and evaluate. SuitShop temporarily turned off AI to improve their help content, making automation far more effective when it came back online.
• Let your best human agents guide your AI. Katie studied her top-performing teammate's tickets to teach AI the best responses and macros.
• Brand voice matters as much as accuracy. SuitShop's AI Agent "Max" is trained to sound warm, helpful, and on-brand. Customers even thank it by name.

The most coachable team member on your support team might not be human.

Brands that want to keep up with rising customer expectations are turning to AI to help meet demand. But as SuitShop’s Director of Customer Experience, Katie Eriks, will tell you, great results don’t come from flipping a switch.

They come from coaching.

Since implementing Gorgias AI Agent, SuitShop has reached a 30% automation rate, all while maintaining a lean CX team and giving every customer the tailored experience they expect (literally and figuratively).

“I consider myself its boss,” said Katie, who runs the entire coaching process solo. With under an hour of weekly maintenance now, SuitShop’s AI Agent runs efficiently, accurately, and on-brand.

Katie spoke at Gorgias Connect 2025 to share exactly how she got there. You can watch her full session below:

The case for coaching your AI Agent 

When brands think about automation, they often imagine flipping a switch and watching repetitive tasks vanish. But in practice, it’s not that simple, at least not if you care about customer experience.

Gorgias encourages brands to treat their AI Agent like a junior teammate — someone you onboard, train, observe, and coach over time.

Brands that do this well are already seeing massive gains:

• 60%+ of customer conversations fully automated
• First response times under 30 seconds
• Consistent CSAT scores of 4.5 and above
• Major cost savings during high-volume seasons

For SuitShop, automation was about creating space for their small team to focus on specialized service. Space to scale without scaling headcount. And space to do it all without losing their voice.

Step-by-step: SuitShop’s AI coaching workflow

Katie and her team had been longtime Gorgias users, but when they turned on AI Agent in August 2023, the results were unremarkable. The responses weren’t inaccurate, but they weren’t helpful enough either.

What Katie learned was to “Be hands-on early. Use downtime to train. And never stop refining.”

So she got to work, not by replacing the tool, but by going deeper into it. Here are her coaching tips: 

One owner, one hour a week

Katie made herself the sole point of contact for training and QA. That might sound like a lot, but over time, it became a light lift.

“At this point, it’s definitely less than one hour per week,” she said. “In the beginning, it was more time-consuming because I needed to create help center articles and Guidance regularly. Now I’ve got it down to a pretty quick thumbs-up, thumbs-down kind of process.”

Katie uses Monday mornings to review AI Agent tickets from over the weekend, when fewer human agents are available and AI takes the lead.

Read more: Why your strategy needs customer service quality assurance

Pause and perfect before scaling

Unlike many retail brands, SuitShop’s busiest time isn’t the holiday rush — it’s wedding season in the summer and fall. So when things quieted down in December, Katie used that time strategically.

She temporarily turned off the AI Agent to regroup.

“I decided to turn it off and really beef up our Help Center,” she explained. “I went back to the tickets I had to answer myself, checked what people were searching in the Help Center, and filled in the gaps.”

She built out content with a mix of blog knowledge, internal macros, and ChatGPT. Once she felt confident the content base was solid, she turned AI back on.

Read more: How to optimize your Help Center for AI Agent

Use data to guide your coaching plan

Once SuitShop’s foundational content was in place, Katie didn’t just sit back and hope for the best. Instead, she built a repeatable feedback loop grounded in data — one that helped her spot opportunities for improvement before they became issues.

Rather than combing through tickets at random, Katie created custom views inside Gorgias to zero in on the most impactful coaching moments:

• Low CSAT tickets: Any conversation that ended with a customer satisfaction score below expectations got flagged. These were clear indicators that something about the tone, accuracy, or clarity of the AI response had fallen short.
• High handover rates: Katie looked at the tickets AI Agent was regularly handing off to humans. Many of these were actually answerable. The handover just meant that guidance was missing, miscategorized, or too vague.
• Agent-tagged tickets: To make this scalable, Katie empowered her team to flag any strange or impressive responses from the AI. By using a tag like AI_agent_feedback, team members could drop tickets into a coaching queue without needing to write a full explanation.

To keep all of this actionable, Katie logs insights in a shared spreadsheet that functions as a live to-do list. Every row includes:

• A link to the ticket
• A summary of the issue
• The resolution (e.g., new Guidance or macro needed)
• A status tracker (not started / in progress / completed)
• A link to the resource she created in response

These insights are also available in Gorgias’s dashboard, where you can identify the top issues customers had.

“Sometimes I do it all in the moment. Other times I’ll log it and come back later when I can take the time to do it right.”

By combining frontline feedback with structured ticket views, Katie turned scattered QA into a consistent coaching system — one that ensures SuitShop’s AI Agent keeps getting smarter every week.

Learn from your human agents

One of Katie’s most effective strategies comes from her own team.

Like many CX leads, she noticed that some agents consistently resolved tickets in a single touch. That pattern, Katie realized, wasn’t just a win for customers, it was a roadmap for an AI-driven support strategy.

Her teammate Tacy quickly became her go-to signal for what the AI Agent needed to learn next.

“I pull her tickets often to see what she’s responded with. It helps AI learn from her directly.”

By reviewing Tacy’s ticket history, Katie identified standard replies that didn’t yet exist as macros or Guidance. If Tacy was writing the same sentence repeatedly or copy-pasting a reply manually, that meant it could (and should) be taught to the AI Agent.

She also tracked Tacy’s macro usage rate. If Tacy frequently used a macro for a certain issue, but other agents weren’t, it flagged an opportunity to standardize responses across the team and the AI.

The key insight? If it only takes one touch for a human to answer, the AI can be trained to do it too.

These small efficiency wins added up quickly, especially during peak season, when the ability to automate just a few extra conversations per day created meaningful breathing room for the rest of the team.

Related: How to automate half of your CX tasks

Make your AI sound human (and on-brand)

Automation without brand voice feels robotic. Katie made sure SuitShop’s AI Agent sounded like a natural extension of the team, and that started with a name: Max.

“We get replies like, ‘Thanks Max!’ from customers who think it’s a real person.”

Using AI Agent’s tone of voice settings, Katie went deep on personalization. She customized everything from sentence structure and greeting format to whether or not emojis and exclamation marks should be used (they shouldn’t, in SuitShop’s case).

Her AI Agent instructions include clear direction on:

• Tone: Warm, empathetic, clear, especially with high-stress wedding-related issues
• Structure: Shorter responses for chat, slightly more detailed for email
• Dos and Don’ts: Specific words or phrases to use or avoid, pulled from real team responses
• First/last name use: Always first name only, to keep things friendly but respectful

Katie also made sure she instructed AI Agent to acknowledge customer emotions — especially frustration — and to offer reassurance when things went wrong.

And because AI responses are written at lightning speed, she regularly reviewed messages to ensure they didn’t come off as cold or abrupt, especially in sensitive situations like delayed wedding orders or size issues close to the event date.

Live coaching: What it looks like in practice

In the workshop, Katie walked through two real support tickets where AI missed the mark and how she used those moments to improve.

In one case, a customer asked a common question: “The navy suit I’m looking at says ‘unfinished pant hem.’ Will the pants need to be hemmed?”

Despite having help articles and macros explaining this exact issue, AI Agent responded: “I don’t have the information to answer your question.”

That was a red flag.

Katie immediately stepped in to coach the agent by:

• Applying the correct existing resources
• Writing an ideal sample reply as an internal note
• Checking tone, empathy, and phrasing
• Testing the fix by pasting the same question into the test environment

“I like to write a short internal note, so if I see that ticket again, I know exactly how I coached it.”

In another case, AI Agent was incorrectly handing off a sizing question about jacket sleeve length. Katie realized that a previous broad handover topic ("sizing and fit questions") was causing confusion by flagging issues that the AI should have been able to handle.

So she deleted the handover topic and replaced it with a clear guidance article — complete with example questions, macros, and links to sizing resources.

“Once I added specific questions in quotes, it made a huge difference.”

What's next: AI tools that help you scale faster

SuitShop didn’t automate 100% of CX — but that’s not the point. At 30% automation (and growing), Katie gives her team more time to specialize, connect, and handle urgent or emotional conversations with care.

Here’s what Gorgias offers to help as well:

• Optimized intent dashboards: These show the most common topics customers ask about, the AI’s performance on those topics, and how much opportunity there is to increase automation.
• Auto-generated knowledge: Based on your macros, Help Center, and even Shopify data, Gorgias can now draft suggested guidance for your review, making it faster to train your AI Agent without starting from scratch.
• Auto QA: This tool scores every AI (and human) response based on resolution, accuracy, communication, and tone. It gives you full visibility across your team and automations, without needing to review each ticket manually.

Whether you’re just getting started or trying to move beyond basic automation, Katie’s approach proves that coached AI outperforms out-of-the-box tools every time.

Want to coach your AI Agent like SuitShop? Book a demo to see how Gorgias can help you scale smarter.

{{lead-magnet-1}}