In arbitrage, people love asking the wrong first question.

They ask: Is this source good?

Usually the better question is: Can I understand what this source is doing to my campaign before it gets expensive?

That difference matters a lot.

A source does not need to be perfect to make money. It needs to be readable. If you can see where the traffic comes from, what kind of user it brings, how it behaves in the funnel, and what part of it is worth protecting, then you have something to work with. If all you get is volume and chaos, you are not really buying traffic. You are renting confusion.

That is why strong buyers do not evaluate sources by hype, CPM, or Telegram noise. They evaluate them by how fast the source reveals the truth.

1. The first thing to check is not price. It is visibility.

Cheap traffic is attractive for obvious reasons. But cheap traffic without visibility is usually expensive in the worst possible way.

A good source should let you break performance down clearly:

  • by placement
  • by creative
  • by device
  • by GEO
  • by time window
  • ideally by zone or publisher-level segment

If the source gives you clicks but hides too much of the traffic shape, every optimization decision becomes slower and less reliable. You may still hit a lucky pocket, but you will have a hard time understanding why it works or why it dies.

That is the difference between a source that gives you traffic and a source that gives you decision-making power.

2. Look at how quickly the source separates good traffic from bad traffic

Some sources feel “bad” in the first hours simply because they need time to stabilize. Others feel “promising” in the first hours because broad cheap traffic produces nice front-end metrics before the quality falls apart.

So the real test is not whether the source looks pretty on day one. The real test is whether the source starts separating useful pockets from useless ones fast enough for you to act.

You want to know:

  • does one placement cluster already look cleaner than the rest
  • does one device segment hold better
  • do some GEOs clearly outperform others
  • does traffic quality drift as soon as volume rises
  • are weak placements easy to identify and cut

A source becomes dangerous when it keeps everything blended together. Mixed traffic can make a campaign look alive while one small profitable segment is quietly covering three bad ones.

3. Judge the source by post-click behavior, not only by click behavior

A lot of bad sources look good at the top of the funnel.

They produce:

  • cheap clicks
  • decent CTR
  • fast movement
  • enough activity to feel “scalable”

Then the funnel tells the truth.

The user clicks, but the intent is weak. Or the registration rate is soft. Or approval falls apart. Or revshare never develops. Or the traffic looks fine until you check approved EPC and realize the source sold you motion, not value.

This is why source evaluation should always include the full path:

  • click-through rate
  • lander click rate
  • conversion rate
  • approved conversion rate
  • approved EPC
  • payout stability

If a source only looks strong before the backend catches up, it is not strong. It is just flattering the front end.

4. Good sources let you control aggression

One of the easiest ways to spot a usable source is to see how it behaves when you change pressure.

Can you increase spend without destroying the traffic shape?

Can you bid up without instantly falling into garbage inventory?

Can you scale one GEO without poisoning the whole setup?

Can you blacklist bad zones and actually improve the campaign?

These are not small details. This is where the source shows whether it is operationally useful or just temporarily cheap.

The best sources are not always the ones with the lowest entry cost. They are often the ones that let you scale in steps without forcing you into a new campaign identity every time you add budget.

5. Watch how the source punishes weak creatives

This sounds strange, but it is one of the most useful signals.

Some sources can hide weak creatives because the traffic is broad enough to still generate movement. Others expose creative weakness immediately. That can actually be a good thing.

A source that punishes weak packaging early is often more honest. It gives faster feedback. It forces better segmentation. It teaches you something real about the campaign.

A source that lets bad creative survive too long can make you feel smarter than you are. That feeling usually ends at scale.

Good arbitrage is easier when the source tells the truth quickly, even if the truth is ugly.

6. Ask whether the source fits your offer, not whether it “works”

This is where many buyers waste the first real budget.

They test a source in the abstract. But sources do not convert in the abstract. They convert inside a specific relationship:

  • this source
  • this audience shape
  • this funnel
  • this payout logic
  • this offer friction

Push traffic may be great for one kind of broad-intent offer and useless for another. Search may be expensive but incredibly clean for transactional intent. Native may work only when the lander does most of the emotional lifting. Messenger traffic may print when the user is warmed up and die when the approach is too direct.

So the source question should never be: Does this source work?

It should be: What kind of offer-funnel combination does this source allow me to run well?

That is a much better question, and it usually leads to better buying.

7. The source is good when it makes optimization clearer, not louder

A lot of bad traffic sources feel exciting. Fast clicks. Big numbers. Constant movement. Enough volatility to make every hour feel important.

That is not always a good sign.

Strong sources usually make your campaign easier to read over time. You understand which pocket is carrying the result. You know what to cut. You know what to protect. When the campaign slips, you have a place to look first.

That is a much better quality than “it can spend.”

Any source can spend. Not every source can teach you where the money actually is.

Bottom line

A good source is not just a place where you can buy traffic. It is a place where the traffic can be interpreted, segmented, and scaled without turning every decision into guesswork.

So before you ask whether the source is cheap, hot, or trending, ask something more useful:

Can this source show me the truth early enough to make smart decisions?

If the answer is yes, you have something worth testing.

If the answer is no, the source may still give you clicks, but it probably will not give you control.

And in arbitrage, control is usually where the margin lives.