The crypto signal industry is a graveyard of half-truths, vanity metrics, and outright fraud. Most services target people who are early in their trading journey and don't yet know what to ask. This guide is the questions to ask — and what the answers should sound like.

You can run this checklist on any service in five minutes.

1. "Can I see your full trade history?"

Right answer: A public page or downloadable CSV listing every closed trade, including the losers, with timestamps that pre-date your visit.

Wrong answer: Screenshots of cherry-picked winners. "DM me for results." A scrollable Telegram channel with auto-deleted messages. Anything that requires you to trust their summary instead of seeing the data.

If they won't show you losses, the losses are either bad enough to scare you off or they're inventing the wins.

2. "What's your win rate AND average risk-to-reward?"

Right answer: Both numbers, together. "63% win rate at 1:2 R:R" tells you something. "We're 80% accurate" alone tells you nothing — you can be 80% right with a 1:0.3 R:R and still lose money.

Wrong answer: Just one number. Vague "we're highly profitable" claims. Refusal to define their methodology.

The math: a strategy with 50% win rate at 1:2 R:R has a +0.5R expectancy per trade. A strategy with 80% win rate at 1:0.25 R:R has a -0.2R expectancy. The high-win-rate one loses money. Without R:R, win rate is meaningless.

3. "Do you publish your stop-loss in advance?"

Right answer: Yes — every signal has the entry zone, stop-loss, and take-profit levels stated upfront, before the trade plays out.

Wrong answer: "We'll tell you when to exit." (No, they won't, because they don't know.) Stops added after the trade went against them. Trailing-stop language with no specifics.

A signal without a pre-published stop is not a trade plan. It's a guess.

4. "What's your worst week / month ever?"

Right answer: A specific number. "Worst week was -8R in March 2026 — three losing days in a row from a momentum-killer event." If they have a verified track record, they know exactly when their worst stretch was.

Wrong answer: "We've never had a losing week." (Either lying or new — both bad.) Vague answers like "we have ups and downs."

Every legitimate strategy has drawdowns. Hiding them means hiding the failure mode that will catch you off-guard.

5. "How does your engine make decisions?"

Right answer: A high-level explanation that doesn't require you to be a quant. Something like "we wait for 4 of 5 technical layers to align: trend, momentum, volume, structure, and volatility regime. We won't enter against the higher timeframe." Specifics enough to verify they have a system, vague enough that competitors can't copy it.

Wrong answer: "Our AI predicts the future." "Proprietary algorithm." "Quantum machine learning." Buzzwords without mechanism = sales copy.

6. "Can I cancel any time?"

Right answer: Yes, monthly billing, cancel from inside your account in two clicks.

Wrong answer: Annual lock-ins with refund refusals. Cancellation requires an email to support that goes unanswered for weeks. "Lifetime" plans for the same price as monthly (the math doesn't work — they're betting you don't notice).

7. "What do I get if I follow your signals and lose money?"

Right answer: "Trading involves risk. Past performance doesn't guarantee future results. We can't guarantee profit. Our edge plays out over many trades, not single ones." This is the legally and ethically honest answer.

Wrong answer: "We guarantee 5x returns." "Money-back if you don't profit." Anything implying their signals can't lose. Both are illegal in most jurisdictions and both are detectable by regulators.

Bonus: the "founder check"

Search the founder's name + "crypto" on Twitter and YouTube. Are they a real human with a track record, or a stock photo with no online history before this service launched?

The serious teams have a public face. The scams use generated headshots from thispersondoesnotexist.com.

What this looks like in practice

Run the 7 questions on any service before you pay. If they can answer 5 or more cleanly, they're probably legitimate. If they dodge more than 2, walk away.

Run the same 7 on TradeVelocity AI:

  1. Full history? Yes — /performance shows every closed trade with a 24-hour delay, sourced from the live database.
  2. Win rate AND R:R? Yes — /performance shows both, plus expectancy.
  3. Stop-loss published in advance? Yes — every signal has entry, SL, TP1/2/3 fixed at publish time.
  4. Worst stretch? Yes — the cumulative R curve on /performance shows every drawdown.
  5. How does the engine work? 5-layer confluence + multi-timeframe + safety vetoes — explained on the homepage.
  6. Cancel any time? Yes — monthly, in-app, two clicks.
  7. Risk acknowledgment? Yes — /disclaimer and bottom of every signal.

You don't have to use TradeVelocity. But use these 7 questions on whoever you're considering. The bad ones will fail.