Short answer
Self-service support is any way a customer solves their own problem without waiting for a person: a help center, an order status page, a returns portal or a chat that can act. It works when the problem gets solved, not when the customer stops asking.- Also called
- Self-service customer service, customer self-service
- Common forms
- Help center, order status page, returns portal, chat that acts
- Fails when
- It answers everyone when the question is about one order
- Measure it by
- Problems resolved and repeat contacts, not deflection
Customers often prefer it. Nobody wants to email a store and wait a day to learn that the parcel is out for delivery. But self-service only helps when the answer it gives is about this customer's situation, and when there is an easy way to reach a person if it is not.
Forms of self-service
- Help center. Articles on shipping, returns, sizing and care, sometimes called a knowledge base. Best for questions with one answer that is the same for everyone.
- Self-service help desk. A helpdesk with a customer-facing side: the help center, a contact form that suggests articles as the customer types, and a page listing their past requests.
- Order status page. Tracking with the latest carrier event and the expected date. It answers "where is my parcel" before it is asked, as long as it keeps updating.
- Chat that acts. An AI agent in the store's chat that reads the shopper's order, answers from the help center, and can cancel, change an address or start a return inside limits the merchant chooses.
- Returns portal. The shopper picks the item and the reason, gets a label and can see where the refund stands.
- Account pages. Order history, saved addresses and subscriptions the customer can change without asking.
When self-service fails customers
- It answers everyone when the question is about one order. A shipping policy is not an answer to "where is #1074?".
- The articles are out of date. The page promises free returns the store stopped offering in spring.
- There is no way out. The chat loops through the same three buttons and hides the option to reach a person.
- It asks for what it already knows, such as an order number the customer is logged in with.
- It ends without fixing anything, and the dashboard still counts the visit as a success.
With Aide Chat on a Shopify store, a shopper logged in to their store account is recognised automatically, so Aide does not ask for an order number it can already see, and their past conversations follow them across devices. A shopper who taps Talk to a person reaches the team at once, without the AI answering first.
How to measure self-service honestly
Two numbers get confused. A deflected contact is one where the customer did not reach an agent after trying self-service. A resolved contact is one where the problem was fixed: the address changed, the return started, the question answered and not asked again. Deflection says nothing about which happened, because a customer who gave up, bought elsewhere or disputed the charge was also deflected.
Measures that keep the count honest:
- Repeat contact. The same customer asking about the same thing within a few days means the first answer did not work.
- Satisfaction on self-service conversations, tracked apart from conversations handled by your team.
- Article signals. Searches that return nothing, and articles that are often followed by a contact.
- Handoff reasons. Why chat conversations reached a person, grouped by reason.
- Contact rate per order. If self-service is working, it falls over time while orders hold steady.
In Aide's analytics, satisfaction is reported for Aide and for the team separately, conversations Aide resolved on its own are counted apart from those a person finished, and every number opens the exact conversations behind it.
Where Aide fits
Aide, the AI platform for customer service, puts self-service inside the chat on Shopify stores. Shoppers can search and read your articles inside the chat or ask instead, and Aide answers from those same articles with the shopper's live order in view, hands the conversation to your team when it cannot help, and shows why each handoff happened.