Custom Services Don’t Fit a Catalog: How Conversational Commerce Turns Open Questions Into Orders

Ecommerce platforms assume the buyer already knows what to purchase: a catalog lists the items, a cart totals them, and checkout takes the payment. That model strains when the product is work whose scope depends on the buyer, such as a renovation, a translation, or a consulting sprint. The buyer arrives with a question, the site answers with a quote form, and the deal goes quiet while somebody prices the job.

Conversational commerce can close that gap when the chat does more than capture a lead. A more useful design lets an AI agent take the open question, turn it into a structured request, answer from current product data, price the custom work in the conversation, and hand the paid order to the professional who delivers it. Here is how that works, with a worked example from a gaming services marketplace that quotes custom orders in chat.

Why Custom Services Stall Before Checkout

Fixed-price stores already lose most carts. Baymard Institute averages 50 published studies and puts the documented online cart abandonment rate at 70.22%. Setting aside shoppers who were only browsing, its survey of US online shoppers finds that 17% abandoned an order over a long or complicated checkout, and 12% left because they could not see or calculate the total cost up front.

Baymard Institute, Cart & Checkout Abandonment Rate Statistics, Baymard Institute, 2026

A custom service starts with both problems built in. The total does not exist until someone scopes the work, usually through a form that asks for details the buyer cannot answer yet, and the quote arrives by email hours or days later.

Speed compounds the damage: research that Harvard Business Review published in March 2011 found that most companies were not responding to online queries nearly fast enough.

James B. Oldroyd, Kristina McElheran, and David Elkington, The Short Life of Online Sales Leads, Harvard Business Review, 2011

For a marketing team, the loss rarely shows up in a dashboard. Paid clicks land on pages without a price, quote requests sit in a shared inbox, and buyers who would have added to the order are never asked.

What a Conversational Commerce Agent Needs to Do

An agent that sells custom work needs more than an FAQ widget does. Look for these capabilities:

• Open entry: The chat starts with the buyer’s words and asks only for the missing details.
• Structured intent: Each conversation becomes a record with fields such as scope, deadline, format and add-ons, which a pricing engine and a delivery team can act on.
• Dated product data: Every fact the agent repeats carries the date it was last checked, and anything still unannounced is labeled that way.
• Quotes in the thread: The agent prices the custom request in the conversation, so the buyer can accept while still on the page.
• Human handoff: After payment, a professional takes over with the agreed brief already written down.
• Thread as order record: Scheduling, progress reports and add-ons stay in the same conversation, so either side can scroll back to what was agreed.

Case Study: Custom Orders in a Gaming Services Marketplace

LFCarry runs an AI gaming companion alongside a marketplace where professional gaming services are optional. Players use the companion to ask questions, compare options, and plan how they play. The company has operated since 2016, and its marketplace covers 70+ games with coaching, carries, boosting, and in-game gold, delivered by vetted professional players it calls Pros.

Game services behave like custom work. Take 2 players who want the same character level: they can differ on starting level, deadline, class, ruleset, daily schedule, and whether they play alongside a Pro or hand the character over. The catalog covers the common paths with calculators and fixed offers. Everything else goes to the conversation. On the company’s Modern Warfare 4 hub, the custom order listing says: Any request, quoted in chat.

Content feeds the same conversation. A leveling-speed guide for World of Warcraft: Forever carries a chat prompt below its FAQ that asks for 2 inputs, the class and the hours a day the player can play, and promises 2 outputs: the week the character reaches level 60 and a price for the part a Pro covers. A reader’s open question becomes a dated plan and a quote, with no form in between.

The company says Smart Chat helps players plan, price, and customize orders and quotes goals the catalog does not list. On its WoW Forever hub page, the chat promises a reply in under 30 seconds.

The product data around the chat carries dates. The same guide shows an update stamp and a What changed log, where an October 25 entry records that the beta closed on October 23. The Modern Warfare 4 hub cites the publisher’s announcement along with the date it was read. Where no date exists yet, the page says so: at the time of writing, the Ranked Play listing noted that the mode had no date and that work would begin when the ladder opened.

Once the buyer pays, a person takes over. The company assigns a Pro who is free to take the work and opens a private chat, in which that Pro settles the exact scope, posts progress, and answers questions while a support team stays on call 24/7. Coaching works the same way, with the player confirming a brief in chat and buying hours.

From that point the order lives in the conversation. Buyer and Pro agree on the start day and the playing hours in the chat, and progress updates arrive there until the work is complete. It is also where the order grows: a WoW Forever player can open with a small level bracket, then add levels, professions, gear, gold, or a 2nd character in the same thread.

Benefits for Marketers and Product Owners

The pattern moves several numbers a growth team already tracks:

• Response time: A reply in seconds meets the buyer on the landing page while intent is fresh, which matters most for paid traffic.
• Fewer abandoned quotes: A price delivered inside the conversation can be accepted before the buyer leaves, so custom requests stop waiting on an inbox.
• 1 thread from question to delivery: Pre-sale questions, scope, payment, progress, and add-ons live in 1 conversation, giving support full context and analytics a clean path from opening message to delivered order.
• Upsell in context: Add-ons offered while an order is running reach a buyer who is already engaged, and the professional on the job can suggest what would help next.
• Content that sells: Guides and help articles that end in a quoting chat turn informational visits into qualified conversations, and the source page travels with the order for attribution.

Audit Your Own Chat Funnel

Before adding another form field or chatbot script, walk a custom request through your own site:

• Time the reply a new chat gets on your highest-intent pages, including nights and weekends.
• Ask for something your catalog does not list and count the steps until a price appears.
• Find out whether your chat can quote, or whether it only collects contact details for a sales queue.
• List the data your agent answers from, and confirm that each fact has a source and a date.
• Check that the person who delivers the work receives the agreed brief without asking the buyer to repeat it.
• See whether the conversation survives checkout as the order record, and whether add-ons are offered there.

Pick the 3 pages that bring you the most custom requests and run this audit there this week. Start with the slowest step: a buyer who stalls on your form has already told you what they want, so answer while they are still on the page.