0
33
When an EA is not doing what you expected, the message you send the seller decides how much help you get. Not because sellers are capricious, but because most reports contain nothing anyone could act on. "It is not working" and "it lost money" are not problems a developer can reproduce, and a developer who cannot reproduce your problem cannot fix it.
Separate the three things that all feel like "it is broken"
There are only three real categories. The EA is not running — no trades at all, or the terminal state is wrong. The EA is running and behaving differently from the description — wrong symbol, wrong hours, far more or far fewer trades. Or the EA is running exactly as described and losing. The first is usually your setup, the second is worth the seller's time, and the third is not a support case at all. Deciding which one you have before you write costs five minutes and changes the entire response.
What a report needs to contain
Terminal build, broker and account type, the exact symbol name including its suffix, the timeframe, and the set file you loaded. Then the specific trades: ticket numbers, times, and what you expected instead. If the complaint is that nothing traded, the Experts and Journal tab entries for the period matter more than anything else, because they usually contain the reason in plain text. A report with those items gets a technical answer. A report without them gets a request for those items, a week later.
Screenshots of the equity curve prove almost nothing
A picture of a falling balance tells the seller that you lost money, which you already said. It does not show what the EA decided or why. The account history export, or even three specific tickets with times, carries more information than any chart image. If you send one screenshot, make it the Journal, not the balance.
Ask a question that can be answered
"Why is this losing?" invites a defensive paragraph. "The description says it trades the London session, but ticket 12345 opened at 02:10 server time — is that expected?" invites a fact. Specific questions get specific answers because they are answerable; broad ones get reassurance because reassurance is the only thing that fits. This is also how you find out quickly whether a seller actually understands their own product.
What refunds are actually for
Marketplace refund policies exist for a product that does not do what it claims, not for a product that traded as described and lost. Knowing that distinction before you buy is what keeps you from treating a purchase as reversible; it mostly is not, and sizing your purchase to that reality is more useful than any argument afterwards. If the product genuinely misrepresents itself, document it with the same specifics as above — that is what a case turns on.
Write the review after a full cycle, not after a bad week
A review written three days in describes the installation, not the strategy. A review written after one losing month describes the market. The reviews that help other buyers, and the ones sellers actually learn from, say how long it ran, on what account, with which settings, and whether behaviour matched the description. That is a review with a fact in it, and there are very few of them.
Our three best-selling EAs: GOLD NEURON (AI) https://www.mql5.com/en/market/product/187329 · ATLAS PORTFOLIO https://www.mql5.com/en/market/product/182751
Our published backtests state period, balance and settings, and our forward accounts are labelled as demo accounts wherever we show them. The full list: fxea365.com/ea/ranking
