A customer reports a bug on Friday and someone has to decide whether to answer it now or pass it to the person who wrote the code. Tiered support is that decision written down: L1 answers first, L2 investigates, L3 fixes. The structure routes volume, so easy requests stop at the cheapest level that can close them.
Tiering earns its keep when a lot of requests arrive and most are simple. A team of five answering its own product has neither, so the levels add a handoff where there was a conversation. What a help desk is covers the category this vocabulary comes from.
What the support tiers are
Most teams that run this model have four or five levels, and the numbering starts below the first human.
| Tier | Who is on it | What it handles | What moves it on |
|---|---|---|---|
| L0 | Nobody. A docs site or a status page | Password resets, how-to questions, order status | The customer finds the answer, or writes in |
| L1 | The first human reply, usually a generalist | Routine questions, known issues with a documented fix | The reply goes out, or the thread goes up with notes |
| L2 | Product specialists with deeper access | Troubleshooting, configuration, reproducing the report | A confirmed bug, or a fix that needs code |
| L3 | Engineers and product experts | Product bugs, data problems, integration failures | The fix ships |
| L4 | An outside supplier | A fault in something you depend on, such as a payment processor | The supplier resolves it, and you tell the customer |
The names are not standard. Some teams fold self-service into L1, and some stop at L3 and call the outside supplier something else. Each level up holds more context, more access and less time.
How work escalates between tiers
Support practice splits escalation into two kinds, as HDI sets out. A functional escalation moves a case sideways to someone with the skill to solve it, and a hierarchical escalation moves it up to a manager or another authority.
| Trigger | Type | Where it goes |
|---|---|---|
| Skills or resources at this level are exhausted | Functional | A specialist at the next level |
| The report needs code-level investigation | Functional | Product or engineering |
| A contract sets a time limit on the answer | Hierarchical | A manager or another authority |
| The impact is critical, or the customer asks | Hierarchical | A manager or another authority |
A handoff works when the context travels with it: the original report, what the last person tried, and the diagnostic output they collected. TechTarget describes the failure mode: without that context the customer explains the problem again, and the wait they notice is the second explanation.
What tiering buys you
Every benefit of tiering is a benefit of volume, and each one has a condition attached.
| What you get | What it needs first |
|---|---|
| A cheaper answer per contact | Enough simple contacts to keep a front line busy |
| Specialist time protected | A queue deep enough to make routing worth a rule |
| The same answer from anyone on the front line | Documented fixes, used consistently |
| Cover across the day | People whose job is the front line |
The arithmetic runs one way. Routing to a cheaper level saves money when that level is full, and it costs a handoff when it is not. The rule that keeps two requests a day off a developer’s desk adds a path for every other request.
Why a team of five should not run tiers
The same guide gives the condition. Tiering pays off when volume is high and most issues are simple, while a team with low volume and consistently complex cases does better with swarming, where several people work the case from the start, according to TechTarget.
| Tiering assumes | A team of five has |
|---|---|
| Volume enough to keep a front line busy | A handful of requests a day, mixed |
| Requests that are mostly simple | Requests that are usually about your own product |
| A separate group of people at each level | The same five people, whoever is free |
| Enough escalations to justify a written rule | Few enough that a name is faster than a status change |
On a team of five the first responder is often the person who can fix the bug, so the L1 and L3 rows describe one person at two points in the day. A handoff is not free even when nobody changes desks: someone writes down the reproduction, the next person reads it before starting, and the customer waits for both. The level label makes that wait visible without shortening it.
Skill difference is the other assumption. Tiering needs abilities that differ enough to route between, and enough escalations that the rule stays familiar. Below that the routing lives in somebody’s head, which is where it was before the levels existed.
Where Chirp fits
Chirp is a shared inbox for small software teams that track their work in Linear. It keeps the escalation as an issue rather than a level: press i in a thread to link it to a Linear issue, create one, or search for an existing one, and the person who fixes it reads the thread it came from.
When every linked issue is Done or Canceled, the thread returns with a reply drafted for you that names the issue. You read it and send it, and no rule in Chirp can send mail.
The Issues view (g i) sorts linked issues by how many customers are waiting, instead of a report per level. It connects to Gmail, Google Workspace or IMAP, sets access per inbox with the roles Admin, Agent and Viewer, and costs $9 per user per month. Shared inbox vs help desk is the category decision that comes before this one.
FAQ
What do L1, L2 and L3 mean in support? They name the levels in a tiered model. L1 is the first human reply, for routine requests with a known answer, L2 investigates what L1 cannot resolve, and L3 can change the code. TechTarget covers the full set, L0 and L4 included.
How many support tiers does a team of five need? One, in effect: whoever picks up the thread owns it until the customer has an answer. Levels start to earn their keep when the front line and the people who can fix things are different people.
Is tiered support the same as swarming? No. Tiering passes a case up one level at a time, while swarming puts several people on it at once. Swarming suits low volume and complex reports, which is the shape most software teams have.
What is tier 0 support? Self-service: the docs site, the knowledge base and the status page. It resolves nothing on its own, and it only works when somebody has written the answer down.
Does a small team need a tiered support structure? Not while the same people answer the email and fix the bug. The structure becomes worth writing down when answering and fixing are separate jobs with separate queues, which describes a support department rather than a team of five.
Next step
Tiering answers how a large support team routes work. On a small one, the question underneath it is where the replies live. What a shared inbox is covers the arrangement that gives one address an owner without adding levels.
