LINE

LINE Number Filtering: 7 Steps

Summary: A complete, step-by-step guide to LINE number filtering: normalize formats, deduplicate, verify registration, segment by activity, and pilot small before scaling.

This LINE Number Filtering: Complete Guide walks you through the whole pipeline of turning a raw, messy contact list into a clean, segmented set of LINE numbers you can actually message — in seven steps, each with a clear "done when" standard so you know the work is truly finished before any campaign goes out.

Why LINE lists need their own filtering routine

LINE is the default messaging app in markets like Japan, Taiwan, and Thailand. It sits at the center of daily communication there — friend groups, customer service, official brand accounts, and even payments all live inside it. For an overseas growth team, that makes LINE one of the highest-value channels in Asia, but also one of the easiest to burn.

Three traits of LINE make filtering especially important:

  • Contacts are a scarce, permission-shaped resource. Unlike platforms where you can reach anyone by username, LINE conversations mostly happen between friends, or from an official account to opted-in followers. A number that never becomes a reachable contact is worth nothing, so every send attempt against a dead number is pure waste and added risk.
  • One person often runs multiple devices or accounts. The same individual may keep a personal account, a work account, and a second line for shopping or community groups. Raw lists routinely contain the duplicates that pattern produces, and duplicates skew every later decision you make.
  • Regional format differences are real. Japan (+81), Taiwan (+886), and Thailand (+66) all store numbers differently — leading zeros, mobile prefixes, and separator styles vary — and some people register by LINE ID or via QR code instead of a phone number at all.

If you skip the filtering stage and send from a raw list, the usual outcome is not just poor delivery. It is higher block and report rates on the accounts you paid good money to prepare. Filtering is not overhead; it is the cheapest insurance in the whole workflow. The same logic applies across channels — teams that also run WhatsApp or Telegram outreach will recognize almost every step below — and if you want to see how filtering fits a multi-channel setup, the platform overview page is a good starting point.

Decide what "filtered" means before you start

Before touching a single row, agree inside the team on what a "filtered" list means. We recommend four layers, in this order:

  1. Reachable — the number or ID actually exists on LINE and can become a contact.
  2. Active — the account shows recent signs of life, not just registration.
  3. Intended — the account plausibly matches your market, language, or use case.
  4. Compliant — the data came from sources you are allowed to use, and people on the list have a way to opt out.

Every step below serves one of these layers. If a step does not improve reachability, activity, fit, or compliance, question why you are doing it.

Four things to prepare before filtering

1. One shared definition for every metric

Write down, in one sentence each, what "valid", "active", and "duplicate" mean for your team. Disagreements here are the number-one reason two people clean the same list and get different results.

2. Test numbers and a control sample

Keep a small set of LINE accounts you control — around ten to twenty — spread across the regions you target. Use them to sanity-check detection results. Also set aside roughly five percent of every list as a control group that you filter but never message; it tells you what happens when people receive nothing, which is a useful baseline when you read campaign numbers later.

3. A compliance note for the list's origin

Record where each list came from and what the person agreed to. A list that performs well but was collected improperly is a liability, not an asset. Filtering is also the last natural checkpoint to remove anyone who opted out.

4. Tooling that keeps records

Choose a filtering setup that logs per-number results and timestamps, not one that just hands you a final "good list". When results look wrong two weeks later, you need the intermediate data. Many teams use a dedicated screening workflow such as social-filter for exactly this reason: the per-number trail is preserved, so a questionable batch can be traced instead of re-purchased.

The 7-step filtering workflow

Step 1: Consolidate every list into one sheet with a source column

Pull all raw lists — purchased files, event sign-ups, past campaign exports, partner referrals — into a single working sheet. Add a source column and a date_added column before you do anything else.

Why first: every later decision benefits from knowing where a number came from. When one batch behaves strangely, you want to isolate it, and the only reliable way is a source tag that survived the merge.

Done when: every row has a non-empty source value, and you can state how many rows each source contributed.

Step 2: Normalize formats — phone numbers, LINE IDs, and country codes

Convert everything to one consistent representation:

  • Strip spaces, dashes, and parentheses from phone numbers.
  • Add the correct country code (+81, +886, +66, and so on) based on the market the list targets, and remove the domestic leading zero where the country's format requires it.
  • Keep LINE IDs in a separate, clearly labeled column — never mix them into phone fields.
  • Where a row might belong to two regions, mark it review instead of guessing.

Done when: a random sample of one hundred rows shows zero format variants, and no phone value sits in an ID column or vice versa.

Step 3: Deduplicate and batch the list

Remove exact duplicates, then near-duplicates: the same number with different formatting from two sources, or one person's two accounts where the entries share other identifying fields. When two duplicates conflict — one row says active, the other says dead — keep the more recent one and flag the pair.

Then split the cleaned list into batches of a few thousand rows. Batching keeps each detection run observable, and if something goes wrong mid-run you only redo one batch, not the whole list.

Done when: the dedupe report shows counts for exact and near duplicates, and every row belongs to exactly one batch.

Step 4: Verify registration and reachability

Now run the actual presence check: does this number or ID correspond to a real LINE account, and can it plausibly be contacted? Run detection in modest, spaced batches rather than one giant blast — aggressive checking patterns create their own problems and can distort results.

Compare detection output against your test numbers first. If known-good numbers fail or known-dead ones pass, stop and fix the detection setup before trusting anything.

Done when: every row carries a registration result, and your test numbers behaved exactly as expected in the run.

Step 5: Segment by activity, language, and region

Registration alone does not make a number worth messaging. Layer the verified list into segments:

  • A — high confidence: reachable and showing recent activity signals, in your target region.
  • B — reachable but quiet: registered, little visible activity. Send to these later, in smaller volumes, with softer messages.
  • C — registered but mismatched: the account exists but the language or region does not fit your offer. Park these; do not force-fit them.

If your list contains language hints (from the source campaign or profile information), attach them now. A Thai-language message sent to a Japanese-language account wastes the number and annoys the person.

Done when: every reachable row is labeled A, B, or C, and each segment's size is roughly what you expected given the list's origin.

Step 6: Prepare accounts and sending rhythm

Filtering output is only half the asset; the sending side must be ready to use it gently. Before any campaign: prepare the accounts you will send from, give new ones a warm-up period of normal human-like usage, assign each account a share of each segment rather than letting one account carry everything, and set daily volume caps per account.

Line up message variants too — at minimum a few openings and body phrasings that rotate — so you are not sending one identical text to thousands of people.

Done when: every segment has assigned accounts, a daily cap, and at least three message variants ready.

Step 7: Pilot small, read the signals, and feed results back

Start with a pilot: one or two small batches from segment A only. Watch delivery, acceptance, blocks, reports, and replies. Compare against your control sample and against your test expectations.

Then decide: if the pilot looks healthy, extend to more of segment A, then carefully into B. If it looks bad, stop — do not "push through" a bad pilot. Feed every outcome back into the working sheet: blocked numbers, opt-outs, and dead-on-arrival entries move to a suppression list so no future campaign ever touches them.

Done when: the pilot has documented results, and the suppression list has been updated from them.

Done-when checklist

StepDeliverableDone when
1 ConsolidateOne working sheetEvery row has source
2 NormalizeStandardized fieldsSample shows zero format errors
3 Dedupe and batchClean batchesDup report written, batches disjoint
4 VerifyRegistration resultsTest numbers behaved as expected
5 SegmentA/B/C labelsEvery reachable row labeled
6 PrepareAccounts, caps, variantsAssignments written down
7 PilotResults and suppression listOutcomes fed back

Five mistakes that ruin a LINE list

  1. Filtering once and never again. Lists age. A number that was reachable in March may be gone in November. Re-filter before every major campaign, not just the first one.
  2. Treating registration as activity. A registered but dormant account will underperform any active one. Keep the A/B/C separation honest; do not flatten it because segment A looks small.
  3. Mixing LINE IDs into phone columns. One mixed field silently corrupts a whole detection run. Keep the columns separate forever.
  4. Filtering without a source column. When one batch underperforms, you will want to know where it came from — and you won't be able to find out.
  5. Ignoring the suppression list. Every contact who opted out or blocked you must be excluded from every future send. Re-messaging people who said no is the fastest route to reports, and reports travel back to the accounts you worked hardest to keep safe.

Where filtering sits in your funnel

Think of the funnel in four stages: data quality, deliverability, response, and conversion. Filtering owns the first stage, and it quietly sets the ceiling on all the others. No copywriting skill fixes a list full of dead numbers, and no sending schedule survives accounts that got burned by messaging them. Teams that treat filtering as a one-time chore usually rediscover this the expensive way; teams that re-filter, segment, and suppress continuously get compounding returns from the same raw data.

The same pipeline transfers almost unchanged to other channels: WhatsApp filtering emphasizes phone-number normalization and anti-ban pacing, while Telegram work leans more on usernames and reachability checks. If you are comparing where to invest next, the WhatsApp channel guide and the Telegram channel guide cover their own versions of these steps. When you are ready to scope tooling and budgets for a multi-channel setup, the pricing page gives a straightforward view of what the plans cover.

Wrapping up

A LINE campaign is only as good as the list underneath it. Consolidate with source tags, normalize formats, deduplicate into batches, verify registration, segment honestly, prepare accounts and rhythms, then pilot small and feed every outcome back. Each step has a "done when" standard — hold the pipeline to it, and the list you send from next month will be measurably better than the one you started with.

FAQ

How often should a LINE contact list be re-filtered?
Re-filter before every major campaign. Lists age quickly as people switch numbers, abandon accounts, or reset devices, so a list that was verified three months ago may contain a meaningful share of unreachable entries by now. Teams that re-filter before each campaign keep delivery quality stable; teams that filter once and reuse the list usually see performance drift downward over time.
What is the difference between a registered LINE number and an active one?
Registration only means an account exists on the platform with that number or ID. Activity means the account shows recent signs of real use. Messaging a registered-but-dormant account still costs you sending volume and increases exposure to blocks and reports, which is why this guide recommends segmenting into A (active), B (quiet), and C (mismatched) layers instead of treating all registered numbers the same.
Can I filter LINE numbers and LINE IDs in the same pass?
You can process them in the same workflow, but never in the same column. Phone numbers and LINE IDs follow different formats and are detected differently, so mixing them into one field silently corrupts detection runs. Keep them in separate, clearly labeled columns and verify each type with its own expected-format check before running any batch.
How large should the first pilot batch be after filtering?
Start with one or two small batches drawn from your A segment only — small enough that a problem does not burn an account, large enough to produce readable signals. Compare results against your control sample before extending. If the pilot looks healthy, widen gradually into more of segment A and then carefully into B; if it looks poor, stop and diagnose rather than pushing through.
Do I need a suppression list if my list came from opt-in sources?
Yes. Even opt-in lists change over time: people opt out later, block accounts, or move on. A suppression list built from opt-outs, blocks, and dead-on-arrival results ensures no future campaign ever re-messages them. Skipping it is one of the fastest ways to accumulate reports, and reports travel back to the sending accounts you invested in preparing.
Contact Us

Need Help? Contact Us Now

Have questions about social media account filtering or data verification? Reach out to our customer service team via Telegram for professional support and real-time assistance.

@appfilter

Click the link above or search @appfilter to contact us

Telegram Customer Service QR Code
Scan to Contact
Scan with Telegram to add us
Telegram