Pre-Sale vs Post-Purchase AI: Why Ecommerce Brands Need Both
The short answer. Pre-sale AI talks to shoppers before they buy: it answers questions about fit, price, shipping and comparison so they can decide. Post-purchase AI talks to customers after they pay: order status, returns, changes and complaints. They are different jobs with different data, different risks and different measures of success. Both sit on one customer journey, and shoppers do not know which side they are on when they ask "will this arrive before the holiday?". If shoppers ask you both kinds of question, both need covering. Whether that means one agent or two depends on how well they share knowledge, how they hand off to people, and how each is priced.
Disclosure: Rep AI publishes this guide and sells an AI agent for pre-sale and for post-purchase conversations. Rep AI is described in the section headed "Where Rep AI fits", with its trade-offs, and is named in the titles of the linked guides. Two Rep AI customers are described in the section headed "Two stores that run both jobs in one assistant", using figures from Rep AI's own published case studies, not from the stores or from an independent source.
This guide describes a method. It does not quote a conversion lift or a ticket reduction as something to expect, because both depend on your store. The two customer examples are one store's result each, as Rep AI reported it in its case study.
Two jobs that look like one
Both jobs happen in the same chat window. They are not the same work.
| Pre-sale AI | Post-purchase AI | |
|---|---|---|
| The shopper is | Deciding | Waiting, or fixing something |
| Typical questions | Does it fit? Is it right for me? Which one should I pick? Will it arrive in time? | Where is my order? How do I return this? Can I change the address? |
| What it reads | Catalog, policies, what the shopper is looking at | Orders, tracking, returns and subscription data |
| What good looks like | The shopper decides with confidence | The problem is solved and the customer does not need to write again |
| The risk if it is wrong | A wrong product claim or a promise you cannot keep | A wrong status, a wrong refund or a customer who feels ignored |
| How to measure it | Compare against a holdout group on revenue per session or orders | Resolution with no human touch and no reopen, repeat contacts, and customer satisfaction |
Because the goals differ, the tests differ. You judge pre-sale AI by whether more shoppers buy than they would have without it. You judge post-purchase AI by whether customers get a correct answer without effort. A single number does not capture both.
Pre-sale AI, briefly
Pre-sale AI engages shoppers before purchase. The useful kind starts conversations from what a shopper is doing, such as lingering on a product page or comparing two items, and answers with your own catalog and policies. It is aimed at the shopper who hesitates and does not open a ticket. For a full definition, see What Is an AI Sales Agent for Ecommerce? How It Works and How It Differs From a Chatbot.
Post-purchase AI, briefly
Post-purchase AI serves customers after they have paid. It reads order and carrier data, answers status questions, handles returns and exchanges, and passes real problems to people. The detail of the common case is in How to Reduce "Where Is My Order" Tickets on Shopify With AI, and the wider picture is in AI Customer Service for Shopify Stores: The Complete Guide to What to Automate and What to Hand Off. This guide does not repeat either.
Where the two meet
The line between them is blurrier than the table suggests.
- Shipping questions before and after. "Will it arrive by Friday?" is a pre-sale question. "Where is it?" is a post-purchase question. Both depend on the same shipping data and the same promise.
- Returns policy. Shoppers read it before they buy. The answer they get at that point shapes whether they buy, and it must match what you tell them after.
- Exchanges and subscriptions. A customer changing an order is still a buyer. The same conversation can lead to a different product or a different plan.
- Repeat purchase. The experience after the first order shapes whether there is a second one.
If one system answers the pre-sale question and a different system answers the post-purchase question, check that their answers agree. A customer who is told one thing before paying and another afterwards remembers it.
What goes wrong when you cover only one side
Support-only AI. A tool built and measured on ticket handling may not be designed to talk to a shopper who has not bought. That shopper may not have opened a ticket. If many of your lost sales come from unanswered pre-purchase questions, a support-only agent may not reach them. You can test whether that is true for your store with a holdout, as described below.
Sales-only AI. A tool built for the moment before purchase may not be designed for what comes after. A shopper who got a quick answer about fit may then ask where the package is and not get one.
One agent doing both badly. A single agent can also go wrong. A customer reporting a damaged item should not be offered an upsell. The agent needs to know whether the person has an order, and a rule that a conversation about a problem does not pitch. Write that rule down and test it.
One agent or two?
The right answer depends on the store. Weigh these factors.
| Factor | Why it matters |
|---|---|
| Shared knowledge | If policies, catalog and tone live in two places, they drift apart. One source of truth is easier to keep consistent |
| Handoff | Whoever answers, a person who takes over should see what was said. Ask how each product passes the conversation on |
| Pricing unit | Pre-sale volume follows your traffic. Post-purchase volume follows your orders. A pricing unit that fits one side may not fit the other, so model both at your own volumes |
| Reporting | Ask whether reporting separates pre-sale conversations from post-purchase ones. Check the answer in the product, not the brochure |
| Your helpdesk | If you already run Gorgias, Zendesk or Kustomer, an agent can sit in front of it. How to Add an AI Agent to Gorgias, Zendesk or Kustomer Without Replacing Your Helpdesk explains the layering |
| Depth in each lane | A product built for one lane may go deeper in it than a product that covers both. Test both on your own questions |
What a public benchmark can and cannot tell you
Gorgias publishes a public AI agent benchmark at evals.gorgias.com and describes it as open-sourced (as captured 2026-10-12). It scores vendors in separate lanes, including one for pre-sale (shopping) conversations and one for support. That structure supports the point of this guide: the two jobs are scored differently.
This guide quotes no scores. Read the benchmark's own page for the current figures, the lane definitions and the method, because scores change as it is refreshed.
How to read a vendor-published benchmark like this:
- Automation is not quality. Check how the benchmark defines automation and how it defines quality. A high score on one does not mean a high score on the other. Look at both.
- Know who publishes it. Gorgias publishes this one and sells a shopping assistant of its own. We cite it because it is public and checkable, not because it settles anything.
- It measures one lane at one time. It does not test your catalog, your policies or your tone. Check how many vendors and conversations a lane covers and how a rank is worked out.
Use a benchmark to choose what to test, and your own questions to decide. For the full comparison with Gorgias, see Rep AI vs Gorgias AI Agent (2026): Keep Gorgias for Tickets, Use Rep to Sell.
How to find out what you need
- Audit your inbox by topic. Sort a sample of recent conversations into pre-sale (fit, shipping to me, policy before buying) and post-purchase (status, returns, changes). The split tells you where your volume is.
- Look for questions that did not become tickets. Read your chat logs, site search terms and abandoned-cart reasons. Pre-sale questions may not reach a helpdesk at all.
- Test pre-sale with a holdout. Show the agent to part of your traffic and not to the rest, then compare revenue per session or orders. Do not compare people who chatted with people who did not: shoppers who start a conversation may already have been more interested.
- Test post-purchase with a resolution definition. Decide what "resolved" means, for example closed with no human touch and not reopened within seven days, and compare before and after.
- Run both for long enough. Include a full buying cycle and any promotion that distorts the result.
Two stores that run both jobs in one assistant
Rep AI's case studies on two stores describe one assistant used for both kinds of question. Each figure below is one store's result as Rep AI reported it, not a rate to expect on your store. The figures are the support-side measure each page reports, and Cabana Life's own words show the pre-sale side. The pages do not define their figures beyond what is quoted here, so do not read one against the other.
Charlie B Collection
Rep AI's case study on Charlie B Collection describes a store that sells apparel and accessories, wholesale and through a Shopify storefront. The page, dated August 24, 2026, says that before Rep AI "every inquiry, no matter how small, turned into a support ticket." The store uses Rep AI for its online inquiries.
Rep AI's case study on Charlie B Collection reports that 94.75% of conversations were resolved by AI with no human needed. The page gives no window for that figure.
Cabana Life
Rep AI's case study on Cabana Life describes a store that sells sun-protective clothing on Shopify. The page, dated September 25, 2026, says that before Rep AI, tickets that arrived on a weekend waited until Monday. The store uses an AI concierge it named Jess.
Rep AI's case study on Cabana Life reports that 89.68% of conversations were resolved by AI with no human needed. The page gives no calendar dates for its reporting period. Jennifer Kovac, Director of Ecommerce at Cabana Life, describes the pre-sale side in the case study:
"The majority of our sales and revenue happen over the weekend. So when a customer had a question, there was a good chance we’d miss the sale because we wouldn’t be able to respond until Monday. As a customer, I wouldn’t wait around on a website for an answer - I’d move on. We needed a way to respond instantly."
Where Rep AI fits
Rep AI has both. Its Sales Agent starts conversations from what a shopper is doing on the page. Its Support Agent lists order status (WISMO), returns and exchanges, cancel and modify requests and subscription edits among its capabilities on hellorep.ai/pricing. Its Concierge plan combines the two. When a conversation needs your team, Rep AI escalates to your team with the conversation attached, including to helpdesks such as Gorgias, Zendesk, Kustomer, Freshdesk and Richpanel. Core plans are priced by monthly website sessions. Rep AI's helpdesk and its email, social and WhatsApp AI are in beta and carry separate pricing, so check their status before you rely on them.
Trade-offs to weigh. Whatever any benchmark says, test Rep AI on your own questions before you rely on it. If you want only a helpdesk with AI for tickets, or only a sales assistant, a product built for one lane may suit you better.
You can try conversations in the simulator at /simulator, or start the 14-day free trial, which needs no credit card.
FAQ
Is Gorgias's shopping assistant good for conversion? Gorgias sells a shopping assistant alongside its helpdesk and lists vendors, itself included, in its public benchmark. Whether any assistant raises conversion on your store is a question for a holdout test on your own traffic. If you are shown a reported lift, ask for the method and the comparison group.
Do I need one AI for sales and another for support? Not necessarily. A store can run one agent for both, or a sales agent next to the AI in its helpdesk. Rep AI's case studies of Charlie B Collection and Cabana Life describe one assistant handling both kinds of question, as single-store accounts. The questions to answer are whether the two share the same knowledge, how conversations pass to a person, and how each is priced as your traffic and orders grow.
What is pre-sale AI for ecommerce? An AI agent that talks to shoppers before they buy, answering questions about fit, price, shipping and comparison, and starting conversations when a shopper shows signs of hesitation. It works from your catalog and policies.
Why do AI chatbots that only deflect tickets leave revenue on the table? A tool measured on ticket deflection may not be designed to talk to a shopper who has not bought, and that shopper may not open a ticket. Whether that costs you sales is a question for a holdout test on your own traffic, not a figure to assume.
Can one agent handle both without mixing them up? It can if it knows whether the person has an order and if you set a rule that a conversation about a problem does not turn into a pitch. Test that rule before launch.
More from REP AI
Shoppers Arriving From ChatGPT, Gemini and Perplexity: How to Convert Them Once They Land
A shopper who arrives from an AI assistant may already have asked a question and seen a short answer, so they may land with a specific expectation and have done part of the research. To convert them: find them in your analytics, make the landing page confirm what the assistant likely told them...
Proactive Chat Messages That Don't Annoy Shoppers: 25 Triggers and Examples for Ecommerce
A good proactive chat message is a short, specific question tied to the page the shopper is on, with one easy next step and a clean exit. Trigger it from a behavior that suggests a question, such as comparing two products or opening a size guide, and not from elapsed time alone. Show it once,...
AI Shopping Assistants for Beauty and Skincare Brands: From Skin Type to Routine
An AI shopping assistant can help customers choose skincare and beauty products by turning what the shopper says about themselves, such as a skin type they describe, a finish they like, a shade they wear or products they already use, into a short list from your own catalog. It should show each...
Want to see what your store would do with Rep?
Run your real Shopify catalog through the simulator before you commit to anything. No sales call required.
