Too Long? Read This in 10 Seconds
Executive bullet points for fast decision-making
- ✓ Labels are per-device and personal: they do not travel between people or survive a phone change.
- ✓ They mark a chat's state but carry no history, ownership or reporting.
- ✓ A label system fails at roughly the point a second person needs to answer the same number.
- ✓ Migrate by mapping labels to two separate ideas: a stage the contact is in, and tags describing what they are.
Labels are the WhatsApp Business app's answer to "which of these forty chats needs something from me". For one person and a manageable list, they work well. Here is exactly where they stop.
What labels do well#
- Colour-coded, filterable, applied in two taps.
- Fast to scan — a colour tells you a chat's state before you read it.
- Free, and available immediately with no setup.
For a solo operator handling a few dozen conversations, a five-label scheme is genuinely enough.
A label scheme that survives a year#
Keep them describing state, not attributes:
🟡 New enquiry
🔵 Quoted
🟠 Awaiting payment
🟢 Order in progress
⚪ Closed
Two rules that keep it usable:
- A chat has exactly one state label. If a chat carries three, the scheme has stopped meaning anything.
- Rename, do not add. Every new label makes the filter list longer and the scheme less consistent.
Attribute labels ("wholesale", "Mumbai", "premium") are where schemes rot. They multiply, and they belong on a contact record, not on a chat.
Where labels break#
The count limit. The app caps custom labels, and the cap is low. Growing businesses hit it and start overloading meanings.
They are per-device. Labels live with the account on the phone. A colleague on a linked device does not get their own workable view, and moving to a new phone risks the scheme entirely.
No ownership. A label cannot say who is handling a chat. With two agents, "Quoted" tells you nothing about who quoted or whether anyone is following up.
No history. You can see a chat is labelled Closed. You cannot see when it moved there, who moved it, or how long it sat in Quoted first.
No reporting. How many enquiries this month, how many became orders, how long the average deal takes — none of it is answerable.
Nothing else attaches. No consent record, no order value, no notes visible to a colleague.
The break point is predictable: labels stop working when a second person needs to answer the same number, or when someone asks a question about last month.
Tags, stages and fields — three ideas labels conflate#
| Concept | Answers | Example |
|---|---|---|
| Stage | Where is this in the process? | Qualifying, Quoted, Won |
| Tag | What kind of contact is this? | wholesale, repeat-buyer, mumbai |
| Field | What specific value do we know? | order value, GST number, preferred slot |
Labels force all three into one flat list, which is why label schemes always sprawl. Separating them makes each one simple: a contact has one stage, several tags, and a set of fields.
Add ownership and history on top — who is assigned, what changed and when — and you can answer questions labels cannot.
Migrating without losing context#
- List every label in use and how many chats carry it.
- Sort each into stage, tag or field. Most turn out to be stages; the rest are tags.
- Design the stage list first. Five or six, matching how you actually sell — see lead generation stages.
- Define a tag vocabulary with a naming rule: lowercase, hyphenated, singular.
repeat-buyer, notRepeat Buyers. - Export what matters before you migrate the number, because chat history does not carry across — exporting chat history.
- Migrate the busiest labels manually, then let everything else be tagged as it comes back into conversation. Do not attempt a full historical backfill; it is rarely worth the hours.
A tagging taxonomy worth copying#
Source: source-ad · source-website · source-qr · source-referral
Segment: wholesale · retail · reseller
Behaviour: repeat-buyer · high-value · price-sensitive
Consent: optin-marketing · optin-updates · opted-out
Operational: needs-invoice · awaiting-docs · escalated
Prefix by dimension, keep the vocabulary closed, and review it quarterly. Tags that nobody has applied in three months should be deleted, not preserved.
Consent tags in particular should not be tags at all once you send campaigns — they belong in structured fields with a timestamp and source, as opt-in done right explains.
When labels are genuinely enough#
One person, under about a hundred active conversations, no reporting requirement, no compliance obligation, no handover risk. That is a real business, and adding a CRM to it is overhead rather than progress.
The moment a second person answers that number, the calculation changes entirely.
Try tags, stages and assignment in AR-Inbox on top of the WhatsApp number you already use.
Related: one number, several agents · building a lead pipeline on chat.
Frequently asked questions
How many labels can I create in WhatsApp Business? +
The app caps the number of custom labels, and the limit is low enough that most growing businesses hit it within a year. Check the current ceiling in the app before designing a scheme around it.
How do I organise WhatsApp Business chats? +
Start with a small set of labels that describe state — new, quoted, waiting on payment, done. Keep them to a handful and rename rather than adding new ones for every variation.
Can labels be shared between team members? +
No. Labels live on the device holding the account, so a second person seeing the same chats through a linked device does not get a separate, manageable view.
What is a WhatsApp CRM tag? +
A tag stored against the contact record in your own system rather than on a phone, so it is shared across the team, reportable, and permanent.