How to Use Whitelists and Blacklists in Ad Networks

Whitelist — a list of placements/sources you explicitly allow. Everything that’s not on the list is blocked by default.

Blacklist — a list of placements/sources you explicitly block. Everything else can run unless you later decide to cut it.

By “placements” I mean for example:

• improve traffic quality;
• protect yourself from fraud and “noise”;
• control conversion cost;
• balance volume and efficiency.

But in reality this is also where things break most often:
too narrow whitelist → no traffic,
too aggressive blacklist → volume crashes, CPA jumps, panic in chats.

Let’s walk through, step by step: when you actually need a whitelist, how not to make it “too tight”, and how to maintain blacklists without panic and losing all your volume.

1. What whitelists and blacklists are (really)

Whitelist — a list of placements/sources you explicitly allow. Everything that’s not on the list is blocked by default.

Blacklist — a list of placements/sources you explicitly block. Everything else can run unless you later decide to cut it.

By “placements” you might track for example:

• website domains;
• apps;
• publishers / sub-identifiers (site_id, placement_id, sub_id);
• bundle id, channel, zone, etc. — depends on the network.

Core idea:

• Whitelist = concentration: keep only the best.
• Blacklist = hygiene: gradually clean out trash and fraud.

2. When you really need a whitelist (and when it only gets in the way)

When a whitelist is a must

1. New offer with strict KPI
   When your budget is limited and your CPA/ROAS target is tough, it’s logical not to start from “the whole network”, but from a proven subset of quality placements.

It makes sense if:

• you already ran in this network and know which sources convert;
• you have historical data for similar offers;
• you have an internal or external list of “top suppliers”.

2. Sensitive brand / strict brand safety
   Think banks, healthcare, government, kids’ products. You cannot afford to show ads on:

• adult websites;
• pirate or shady sites;
• political/extremist content, etc.

In this case it’s easier to pre-approve a whitelist of domains/categories than to fight a reputation crisis afterwards.

3. Working with premium, limited inventory
   For example:

• premium news media;
• major publishers;
• high-quality in-app inventory.

Here you consciously sacrifice reach, but bet on quality.

When a whitelist does more harm than good

1. You’re new to the network and have no data
   If you just entered a new ad network and have zero historical performance, a tight whitelist often means: “we’re targeting blind, and also limiting ourselves”.

In that case it’s better to:

• start a bit wider (no whitelist);
• collect initial stats;
• build the first whitelist from real data later.

2. Small budget / soft KPI
   If you don’t have ultra-strict CPA/ROI goals and your mission is “get volume and learn”, a whitelist only cuts your reach. These campaigns are easier to run with a blacklist-first approach.

3. How not to suffocate your whitelist

Classic mistake: you take 10–20 “favorite” placements, turn them into a whitelist — and volume doesn’t grow. Bids are high, but impressions are still tiny.

Principle 1. Start broader, tighten later

If your brand and offer allow it, use this logic:

1. Start “moderately broad” (no whitelist, only basic blacklist by obvious bad categories).
2. Collect data for at least:
   • 3–7 days with stable traffic, or
   • N conversions (e.g. 100+ leads or 200+ installs).
3. Select top sources by:
   • CR (conversion rate);
   • CPA/CPI;
   • quality metrics (retention, approvals, ROI, etc.).
4. Build your first whitelist from these real winners.

This way your whitelist is based on numbers, not on random guesses.

Principle 2. Don’t make whitelists out of 5–10 placements

As a rule of thumb: a whitelist of 5–10 placements is not a whitelist, it’s a choke point.

Rough guideline:

• For at least somewhat stable volume in a big network, you usually want:
  • 50+ placements minimum for a whitelist start;
  • ideally 100+ as you expand.

If you see that:

• you’re already bidding aggressively on your current whitelist, and
• still can’t get enough volume —

→ the answer is expand the whitelist, not squeeze it further.

Principle 3. Split your whitelist into tiers

Instead of one big “good list”, it’s much easier to manage a tiered system:

• Tier 1 (Core whitelist)
  Best of the best: top CR, low CPA, strong quality.
  For these you can:
  • run a separate campaign;
  • bid higher;
  • prioritize budget.
• Tier 2
  Solid, middle performers: acceptable KPI, decent volume.
  This is your main scaling pool.
• Tier 3 (Test pool)
  New or uncertain sources with potential.
  They:
  • run with capped bids/budgets;
  • can graduate into Tier 2/1 if they prove themselves.

Your whitelist becomes a living system, not a static “holy list”.

4. Blacklists: how to update them without panic and killing volume

The other extreme: campaign misses KPI → someone checks the report → sees “suspicious” placements → drops dozens/hundreds of sources into the blacklist → traffic collapses, algorithms break, chat is on fire.

What not to do

• Don’t cut everything that didn’t convert for 2 days.
  Some placements:
  • have slower conversion cycles;
  • convert with a delay;
  • depend on time of day or weekday.
• Don’t blacklist after 1–2 clicks or a handful of impressions.
  You need minimum volume before you decide a placement is truly bad.

A more systematic way to handle blacklists

Step 1. Define thresholds for decisions

For example:

• minimum X clicks (or impressions) before evaluation;
• minimum Y spend per placement;
• observation window (e.g. 3–7 days).

If a placement hasn’t hit these thresholds yet, it’s too early to blacklist it.

Step 2. Separate reasons for blocking

1. Hard reasons (block immediately):
   • obvious fraud (bots, click flooding, abnormal patterns);
   • brand safety violations;
   • prohibited content (adult, politics, violence — if critical for your brand).
2. Soft reasons (block based on performance):
   • CPA/CPI much higher than campaign average;
   • very low CR with enough volume;
   • bad post-conversion quality (refunds, churn, rejection rates).

Here you act like this: enough volume + consistently bad KPI → candidate for blacklist or downgrade.

Instead of blacklisting — “downrank” the source

Sometimes a placement:

• is not great,
• but also not a total disaster.

Instead of putting it straight into the blacklist, you can:

• lower its bid (if the network allows bid by site/sub_id);
• move it into a low-priority campaign with smaller budget.

You don’t kill volume completely, but you stop weak sources from eating the main budget.

5. How often to update white/blacklists

The key is: lists should move, but not twitch.
If you constantly re-write them every few hours, optimization algorithms never stabilize.

Reasonable rhythms:

• Active testing / new campaign:
  • review placements every 1–2 days;
  • big structural changes – no more than every 3–4 days.
• Stable setup:
  • review 1–2 times per week;
  • urgent changes only for fraud or hard violations.

Main idea: lists are dynamic, but not hysterical.

6. Practical checklist: how not to lose volume

Before you turn on a whitelist, check:

1. Do you have enough candidates (50+ placements, ideally 100+)?
2. Have you checked them by:
   • conversion rate;
   • CPA/CPI;
   • quality metrics (LTV, approvals, retention, etc.)?
3. Are you not over-restricting:
   • geo;
   • device;
   • time of day?
4. Do you have a plan for expanding the whitelist:
   • test campaigns;
   • Tier 2/3 test pool?

Before you send a placement to the blacklist, ask yourself:

1. Do I have enough data on it:
   • more than N clicks;
   • more than M spend;
   • reasonable observation window?
2. Am I sure the issue is the placement, not:
   • creative mismatch;
   • landing page problems (slow, broken, not mobile-friendly);
   • wrong targeting/segment?
3. Is this a hard reason (fraud, brand violation) or economic (too expensive)?
   • hard reason → blacklist now;
   • economic → maybe try lower bids first.

7. Common mistakes and how to avoid them

Mistake 1. “Favorite placements” whitelist forever

Scenario:
you once built a list of winners (for a past offer), turned it into a whitelist — and never touched it again.

Problem:

• markets change;
• audiences burn out;
• publishers change their own monetization.

Fix:
re-check your whitelist every 1–2 months,
add new winners from test pools,
remove sources that consistently miss KPI.

Mistake 2. Panic-driven massive blacklist

“Yesterday KPI dropped → today everything that didn’t convert goes to blacklist.”

Problem:

• you confuse algorithms;
• you block learning;
• you destroy your own reach.

Fix:

• define clear thresholds;
• separate hard vs soft reasons;
• don’t make global decisions based on a single bad day.

Mistake 3. No lists at all

“Why bother with whitelists/blacklists? Algorithms are smart.”

Over time:

• weird sources slip in;
• you pay for a lot of junk;
• fraud and low-quality placements accumulate.

Fix:

• at least maintain a basic blacklist: known fraud/bad placements;
• review periodically;
• as you collect data, gradually move to more granular list management.

8. How to organize work with teams and partners

If you’re not working alone, you really don’t want whitelists/blacklists living in someone’s private file that no one else can see.

Good practice:

1. Single source of truth
   A shared spreadsheet / CRM / internal tool where the current whitelist/blacklist for each network/offer lives.
2. Clear update rules
   Who is allowed to make changes, conditions for adding/removing placements, and minimum stats required to take action.
3. Change log
   When the placement was added/removed, why, and who requested the change.

That’s how you avoid “someone blacklisted something, volume dropped, and nobody knows why.”

9. Example table template for whitelists & blacklists

Below is a simple table template you can copy into Google Sheets / Excel / Notion and use as your master file for managing placements across campaigns and networks.

Master placements table (example)

Network	Offer	List type (WL/BL)	Tier (for WL)	Source ID (site_id / sub_id / bundle)	Domain / App name	Status (active / paused / removed)	Reason (performance / fraud / brand / test)	KPI snapshot (CR, CPA, ROI)	Date added	Added by	Notes / context
Network A	Loan EU	WL	Tier 1	12345	example.com	active	performance	CR 4.2%, CPA $8 (target $10)	2025-11-10	Alex	Top source for DE & FR, prioritize
Network A	Loan EU	BL	–	98765	shady-site.net	active	brand	–	2025-11-12	Maria	Adult content, blocked immediately
Network B	App US	WL	Tier 2	app_567	Example App	active	performance	CR 1.8%, CPI $1.4 (target $1.5)	2025-11-15	John	Stable volume, average quality
Network B	App US	BL	–	4444	–	active	performance	0 conv / 150 clicks / $60 spend	2025-11-18	John	Over 3x target CPI, moved to BL

How to use this template in practice

• Filter by network/offer to see active whitelists/blacklists per setup.
• Filter by list type (WL/BL) when you want to export lists to an ad network or share with a partner.
• Filter by “Tier” to quickly see your best Tier 1 sources and your test pool.
• Use “Reason” + “KPI snapshot” to understand why something got onto a list (no more “who did this and why?”).
• Use “Date added” to review older decisions and possibly re-test some placements later if conditions change.

10. Conclusion: treat lists as a living system, not a one-time setting

Whitelists and blacklists are not just checkboxes in the UI. They’re core tools for:

• controlling traffic quality;
• protecting your brand;
• managing CPA/ROI;
• scaling without chaos.

Key ideas to take away:

• Whitelist = focus on the best traffic and brand safety.
• Blacklist = keep inventory clean and economics healthy.
• Don’t make whitelists tiny and static.
• Don’t turn blacklist management into a daily panic.
• Base decisions on data, not vibes from a single bad day.

If you build even a simple, consistent process around whitelists/blacklists, they stop being a source of stress and become one of your strongest levers for performance and control.