GoldenrozBook a call

Ecommerce chatbot development for order status, returns and product questions

An ecommerce chatbot answers shoppers on your store, WhatsApp or social inboxes: where an order is, how to return an item and which product fits, using live data from your store platform. Goldenroz builds it with order lookup, returns flows, catalog answers, opt-in cart reminders and handoff to your support team. Projects start at $25,000, fixed price per phase.

From $25,000, fixed price per phase

Updated 2026-11-14. By Shai Goldenberg

What usually goes wrong

  • Where is my order
    Order status questions make up a large part of most stores' support volume. Each one means an agent opening the store admin and the carrier's site to copy a tracking line.
  • Peak season queue
    Volume multiplies around Black Friday and the December holidays. Temporary agents take weeks to train and reply times stretch just when shoppers are deciding.
  • Returns by email thread
    A return takes four emails: the request, the order number, the reason and the label. Customers give up or open a payment dispute instead.
  • A widget that knows nothing
    The packaged bot answers from a short FAQ list. It cannot see the order, the stock level or the return window, so it sends every real question to a form.

What is included

  • Order status and changes
    The bot verifies the shopper, reads the order and the carrier tracking, and explains the status in plain words.
  • Returns and exchanges
    Eligibility checked against your policy and the order date, the reason collected, and the label or return request created in your returns system.
  • Catalog answers
    Product questions answered from your catalog data, size guides and compatibility tables, with live price and stock. If the data does not say, the bot does not guess.
  • Store platform integration
    Connection to Shopify or WooCommerce through their official APIs, plus your helpdesk, returns tool and shipping provider.
  • Opt-in cart and back-in-stock messages
    Reminders sent only to shoppers who agreed to receive them, on the channel they chose, with a clear way to stop.
  • Handoff, fallbacks and cost log
    Transfer to your agents with transcript and order attached, fixed replies when a model fails, and cost per conversation recorded.

What a chatbot for ecommerce should handle and what it should hand over

The bot is worth building for requests that are frequent, factual and tied to data you already hold. Order status, delivery estimates, return eligibility, product facts and store policies fit. Policy stays outside the model: return windows and refund limits come from your systems.

Other requests go to people by rule: damaged goods with photos, refunds above an amount you set, suspected fraud, wholesale inquiries and any shopper who asks for an agent. The agent receives the transcript, the order and a short summary, and the shopper does not repeat anything. A wider support program across channels is described on the customer service chatbot page.

Shopify chatbot: apps from the store or a custom build

The Shopify App Store has many chat apps, and for a store with standard questions one of them is the right purchase. A custom build makes sense when the bot must reach systems beyond the store, such as an ERP, a warehouse system or a subscription platform, when you sell on WhatsApp as well as the site, or when per-conversation fees have grown with your volume.

The integration uses Shopify's Admin and Storefront APIs with the narrowest access scopes that work, and WooCommerce's REST API on that platform. The bot can read only the orders of the shopper it has verified, by a signed-in session or by order number plus a second detail. Store work beyond the bot, such as custom apps and migrations, sits under Shopify development.

On WhatsApp, messages the store starts, including cart reminders, must use templates Meta approves, and Meta charges for them by category. Check Meta's current rate card for amounts. Consent rules for promotional messages differ by country and are your counsel's call. We build the opt-in record and the stop mechanism.

Building for peak season load

A bot that works in September can fail in late November. Store APIs enforce rate limits, model providers slow down under load and a carrier's tracking service may time out. We queue requests, cache what is safe to cache, set strict time limits on each model call and keep fixed replies ready for when nothing responds.

New intents are frozen before the peak and the test set of real conversations runs before any release. The process and typical timelines are on the AI chatbot development page, and a build meant for the holidays should start well ahead of them.

How the work runs

  1. Scoping call

    Before the project

    We review real transcripts, pick the intents worth automating and agree what the bot must never do. You get a fixed price.

  2. Prototype

    About 1 to 2 weeks

    A working bot on the top few intents, tested against your past conversations in a private channel.

  3. Production build

    Roughly 4 to 8 weeks

    System integrations, fallbacks, handoff, logging and channel setup. The range depends on integrations and on WhatsApp account approval.

  4. Limited launch

    At launch

    The bot handles a share of conversations or selected hours first. Your team reviews transcripts daily.

  5. Widen

    After launch

    New intents are added in order of volume, each tested against the conversation set.

What it costs

Each phase has a fixed price agreed before work starts. Model usage and WhatsApp messaging are billed by the providers to your own accounts.

See full pricing

  • Support or ordering bot
    Fixed price per phase
    From $25,000
  • Multi-channel or multi-brand platforms
    Scoped per project
    $100,000 and up
  • Scoping call
    Ends with a written scope and a price
    $0
  • Model and WhatsApp usage
    Billed to your own accounts
    Paid to the providers

Who it is not for

  • Stores with budgets under $25,000. A chat app from your platform's app store is the better fit.
  • Unsolicited promotional messaging to purchased or scraped lists.

Who does the work

CTO and technology partner

Shai leads the technology side: business software, SaaS and mobile apps, payments and escrow systems, Postgres security, AI agents built with Claude Code and MCP, and IT and infrastructure projects with senior infrastructure engineers.

Questions buyers ask

How much does an ecommerce chatbot cost?

Projects start at $25,000 at a fixed price per phase, agreed after a free one-hour scoping call. Model usage and any messaging fees are separate and paid from your own accounts.

Does it work with Shopify and WooCommerce?

Yes. The bot connects through each platform's official API. For other platforms we check what the API exposes during scoping, since the bot can only answer from data it can read.

Can the bot recommend products?

It can narrow the catalog by what the shopper says they need and explain differences from your product data. It does not invent specifications or promise fit where your data is silent, and it shows live price and stock.

Can it send abandoned cart messages?

Yes, to shoppers who opted in, on the channel they agreed to. Each message carries a way to stop.

Should we buy a chat app or build?

Buy when your questions are standard and your data lives in the store. Build when the bot must act in several systems, follow your own rules, run on WhatsApp and web together, or when usage-based fees have become significant.

Guides on this topic

Talk to the person who will do the work

One hour, no sales team. You talk to Shai Goldenberg, the CTO, about your systems or your product, and you leave with a clear next step and a price range.

Or send the details

Book a 1-hour consultationWhatsApp us