Not all support tickets are equal. A buyer asking “where is my order?” on Amazon at hour 23 of the 24-hour response window is not the same as a pre-purchase question about shipping costs on your Shopify store. The first is a policy violation if unanswered. The second can wait ten minutes. The problem is that most support inboxes treat both tickets identically: they arrive in sequence and get answered in sequence. A ticket prioritization framework changes that. This guide covers a four-tier triage system built specifically for marketplace sellers, and how to configure it in your helpdesk.
At a Glance
- On Amazon, an unanswered message at hour 24 is a metric breach, not just a delay.
- Four tiers: P1 SLA-breach risk, P2 open dispute or feedback, P3 standard query, P4 non-urgent.
- P-tier determines response window, not queue position.
- eDesk’s SLA timer and Ava AI sentiment analysis apply this triage automatically.
- Amazon and eBay have different P1 criteria. Same tier, different triggers.
- A triage system converts prioritization from a judgment call into a repeatable rule.
Why Marketplace Sellers Need a Specific Triage Framework
Generic ticket triage frameworks sort by urgency, complexity, or customer value. Marketplace triage has a fourth dimension: platform compliance risk. An Amazon message that goes unanswered for 24 hours is not just a delayed reply. It counts against your response rate metric, which affects account health and Buy Box eligibility. An eBay dispute that you do not respond to within the platform’s window can result in an automatic finding against you.
The marketplace compliance layer changes the P1 definition. On your own website, P1 might be a high-value order with a complaint. On Amazon, P1 is any message approaching the 24-hour window, regardless of what the message contains. A trivial “thanks for your order” message from a buyer is still P1 if it is approaching the SLA limit and you have not replied.
The 4-Tier Marketplace Ticket Prioritization Matrix
The table below is the framework. Each tier has a defined trigger, a response window, and the account risk it carries. Apply this as your team’s standing triage policy.
| Tier | Response window | Account risk | Trigger criteria | Example tickets |
| P1 | Immediate | SLA breach; ODR impact | Amazon/eBay message within 1h of SLA window closing; A-to-Z claim notification | “Where is my order?” at hour 23 on Amazon; eBay case about to escalate |
| P2 | Within 1 hour | A-to-Z, feedback, dispute | A-to-Z claim opened; negative feedback left; open eBay Return Centre case | A-to-Z claim notification in Seller Central; eBay dispute opened |
| P3 | Within 6 hours | Buyer dissatisfaction risk | Standard post-purchase query; complaint without open dispute; return request | “How do I return this?”; damaged item report; wrong item received |
| P4 | Within 24+ hours | Minimal if SLA met | Non-urgent, no account health impact; pre-purchase on own store; general enquiry | “Do you ship to Canada?”; newsletter unsubscribe query; product compatibility question (own store) |
What Each Tier Means in Practice
P1 SLA-breach risk Response window: Immediate | Trigger: Amazon message within 1 hour of the 24-hour window closing; A-to-Z claim just opened
P1 is defined by time pressure, not message complexity. An Amazon message that arrived 23 hours ago is P1 whether it says “where is my order?” or “great service.” The compliance clock is the same regardless of content. In your helpdesk, P1 tickets should be visually distinct from everything else. Red in eDesk, pushed to the top of the queue, with an agent alert. Agents should have authority to stop what they are doing and respond to a P1 ticket immediately.
P2 Open dispute or platform action Response window: Within 1 hour | Trigger: A-to-Z claim opened; negative feedback left; eBay case escalated to platform review
P2 is where the marketplace has already moved from a buyer message to a platform process. An A-to-Z claim notification is P2. A negative feedback post is P2. These are not SLA risks in the same sense as P1, but they are account health risks that require a fast, considered response. The difference from P1 is that P2 tickets require a human decision. You cannot auto-respond to a P2 ticket the way Ava handles a WISMO query.
P3 Standard post-purchase query Response window: Within 6 hours | Trigger: Return requests, complaints without open disputes, damaged item reports, wrong-item queries
P3 is the bulk of your daily queue: the operational work of ecommerce support. These tickets require a response but do not carry immediate account health risk. A six-hour internal SLA is appropriate for most P3 ticket types. For teams with Ava handling WISMO automatically, P3 is where agent time concentrates.
P4 Non-urgent Response window: 24+ hours | Trigger: Pre-purchase questions on own store, non-urgent general queries, informational requests
P4 covers contacts that need a response but carry no account health risk and no urgency signal. Pre-purchase questions on your Shopify store during off-peak hours, product compatibility requests that require research, and general informational queries can wait 24 hours without any negative consequence. P4 does not mean ignored. It means scheduled: worked through in a dedicated off-peak session rather than interrupting higher-priority responses.
How to Configure This Framework in eDesk
- Set SLA timers per channel to define P1 automatically. In eDesk, set a separate SLA timer for each marketplace channel. For Amazon, configure the SLA window at 20 hours and set an amber alert at 18 hours and a red alert (P1 visual) at 22 hours. When a ticket turns red in eDesk, agents see immediately that it is P1 without having to calculate the timestamp manually. This converts P1 identification from a judgment call into an automatic visual signal.
- Route P2 dispute notifications to a priority queue. A-to-Z claim notifications and eBay case escalations arrive as specific message types. In eDesk, create a routing rule that places these message types in a dedicated P2 queue and assigns them to a senior agent or team lead automatically. P2 tickets require human decisions. They should not sit in the general queue where they compete with P3 volume for agent attention.
- Let Ava handle the P3 volume. The majority of P3 ticket types, WISMO, standard return requests, delivery status queries, are automatable. In eDesk, configure Ava to handle these ticket types using live order data from your connected channels. A P3 ticket that Ava resolves automatically is one fewer item in the P3 queue, which means agents have more time for the P2 decisions and the P3 tickets that do require a human response.
See how eDesk’s SLA timers and Ava triage work together in one helpdesk. Explore eDesk triage features
- Define your P4 process and schedule it separately. P4 tickets should not be mixed with P1, P2, and P3 in the same queue. Create a dedicated P4 label or folder in eDesk and route non-urgent contacts there. Assign one agent a scheduled P4 session each day (30 to 60 minutes at a low-traffic time) to clear the P4 backlog. This prevents P4 tickets from accumulating while also preventing them from interrupting higher-priority work.
- Review your triage rules monthly. Triage criteria drift over time if you do not review them. A ticket type that was P3 when you had one marketplace may become P1 when you add Amazon. A routing rule that worked at 200 tickets per month may need adjusting at 2,000. Once a month, pull eDesk’s SLA compliance report by channel, check whether any ticket types are consistently reaching agents outside their target window, and adjust the routing rules or SLA timer settings accordingly.
What Automatic Triage Changes
eDesk’s AI agent Ava classifies incoming messages across 20+ query types with 95%+ accuracy. When Ava identifies a message, it routes it based on ticket type and channel before an agent reads it. For marketplace sellers, this means P1 Amazon messages turn red in the queue automatically and WISMO queries (the bulk of P3 volume) route to Ava for auto-resolution, without the agent having to decide what to do with each ticket first. The judgment cost of prioritization, which adds seconds per ticket across hundreds of tickets per day, is removed from the agent’s workflow.
What This Looks Like in Practice
A support manager for a seller operating on Amazon, eBay, and Shopify was finding that their team regularly missed Amazon SLAs despite not being overloaded. The root cause: Amazon messages were mixed in the general queue with Shopify contact-form submissions that had arrived earlier but had no SLA pressure. After setting up eDesk’s SLA timers per channel, Amazon messages turned red at 22 hours and appeared at the top of the queue automatically. The team’s Amazon response rate moved from 93% to 99% within two weeks, without any additional staffing or process change beyond the timer configuration.
Key Takeaways
- Not all tickets are equal: on marketplaces, compliance risk defines P1, not message complexity or customer value.
- The four tiers are P1 (SLA-breach risk), P2 (open dispute), P3 (standard post-purchase), P4 (non-urgent). Each has a defined response window.
- eDesk’s SLA timers apply P1 and P2 visual signals automatically, converting prioritization from a judgment call into a system.
- Ava auto-handles P3 WISMO volume, freeing agents for P2 decisions and the P3 tickets that require a human response.
- Review triage rules monthly. Criteria that worked at one volume or channel mix will need adjusting as your operation grows.
FAQs
How do I decide which customer service tickets to answer first?
On marketplaces, start with compliance risk: Amazon messages within one hour of the 24-hour response window are always first, regardless of what they contain. Then open disputes and negative feedback (P2), which require human decisions and carry account health risk. Then standard post-purchase queries (P3) within six hours. Non-urgent, account-health-neutral contacts (P4) can wait 24+ hours. Apply this as a standing rule rather than a per-ticket judgment.
What is ticket triage?
Ticket triage is the process of sorting incoming support contacts by priority before assigning them to agents. Rather than answering tickets in the order they arrived, triage routes each ticket to the right priority tier based on defined criteria: SLA risk, account health impact, query type, and urgency. For marketplace sellers, triage must account for platform-specific compliance windows that override standard customer-service urgency rules.
How do I set ticket priority rules in my helpdesk?
In eDesk, configure SLA timers per marketplace channel so tickets that approach the response deadline turn amber and then red automatically. Create routing rules that place specific message types, such as A-to-Z claim notifications, into a dedicated P2 queue and assign them to a senior agent. Enable Ava to auto-resolve P3 WISMO queries using live order data. Create a P4 label for non-urgent contacts and schedule a daily session to clear it separately from your main queue.
Related: how to set up customer service SLAs | how to automate ecommerce customer support
Ready to apply a 4-tier triage framework automatically in one helpdesk? Start your free 14-day trial
