E-commerce
June 26, 2026
An e-commerce knowledge base powers agents, the chatbot, the FAQ, and sometimes the help center. If it is disorganized, answers become contradictory: a delay on one page, another rule in a macro, a forgotten exception.
The structure must help find the right answer, identify the source of truth, and update rules at the right time. The customer benefits from more consistent support.
This guide shows how to structure an e-commerce knowledge base with concrete examples.
Summary
Why does structure matter?
A knowledge base is not a folder filled with texts. It is the source that allows support to answer reliably. If it is difficult to navigate, agents improvise and the chatbot retrieves weak information.
The structure must make the correct answer visible, along with its validity date and its owner.
A reliable base reduces contradictions even before the chatbot responds.
Convert over 2,000 customers on average per month with Qstomy.
The world’s 1st Shopify AI dedicated to customer conversion
Empowering 200+ e-commerce merchants
Which categories should be created?
The basic categories are orders, delivery, returns, refunds, payment, account, product, warranty, subscription, personal data, promotions, and incidents.
Each category must be linked to customer motives and not just to internal teams.
How to manage sources of truth?
Each response must have a source: published policy, order system, logistics rule, product sheet, legal, finance, or validated support decision. This source must be visible to avoid invented answers.
When a rule changes, the database must indicate which pages, macros, and chatbot responses need to be updated.
How to write internal articles?
An internal article must explain the customer response, limits, proof to be requested, cases to be transferred, and known exceptions. It must be precise enough to act upon, but not filled with unnecessary jargon.
The customer-facing version can be shorter, but it must remain consistent with the internal one.
How do I connect the chatbot?
The chatbot must access validated sources, not old or contradictory notes. The content must be structured by intent so that the bot retrieves the correct rule.
It is also necessary to plan for cases where the bot should not answer alone: payment, data, disputes, or exceptions.
This governance prevents an old internal article from contradicting a public page or a specific exception from being repeated as a general rule.
Which flow to follow?
The flow must organize knowledge by usage.
Classify content by customer intent, support category, source, and risk level.
Assign an owner, a review date, and a source of truth to each rule.
Write customer response, agent procedure, evidence, limits, exceptions, and escalations.
Connect FAQ, help center, chatbot, and macros to the same validated database.
Measure errors, contradictions, obsolete articles, and unanswered questions.
Which examples should be used?
A "damaged product return" article can contain photos to request, deadline, possible decision and quality transfer. A "double charge" article must distinguish between authorization, collection and refund.
A "promo code" article must specify non-accumulation, country, date and evidence to collect in case of contradiction.
When to revise?
The database must be revised after any change in policy, carrier, product, payment method, country, campaign, or incident. Recurring tickets can also signal an obsolete source.
An unmaintained database quickly becomes a risk for providing incorrect answers.
Which KPIs should be monitored?
Track viewed articles, corrected errors, contradictions, obsolete content, sourceless answers, agent satisfaction, customer satisfaction, and transfers related to an incomplete base.
This data shows whether the knowledge is truly usable.
Which mistakes should be avoided?
Avoid mixing customer content with internal notes, failing to designate an owner, keeping old macros, or letting the chatbot read unverified sources.
The database must remain clean, governed, and easy to maintain.
How can Qstomy help?
Qstomy can connect the chatbot to tickets, FAQs, help centers, knowledge bases, NPS surveys, orders, product content, and escalation rules to provide clear answers, and then transfer sensitive cases with an actionable summary.
The chatbot helps the customer move forward without inventing a strategy, a FAQ answer, a support rule, an NPS measurement, or a knowledge source that has yet to be confirmed by a team or reliable data.
Explore AI support, the AI sales agent or request a demo.
Key takeaways
Key Takeaways
A knowledge base must structure categories, sources, owners, evidence, limits, exceptions, and escalations.
What the Customer Needs to Understand
The customer must receive the same rule regardless of whether it comes from the FAQ, the chatbot, or an agent's response.
The Right Boundary for the Chatbot
The chatbot can leverage the knowledge base, but only if it is validated, maintained, and segmented by intent.
Enzo
June 26, 2026
