What actually changes the development scope
Clear indicator or price-action rules are faster to implement than discretionary descriptions that first need to be converted into deterministic conditions.
Lot sizing, exposure limits, daily limits, emergency protection and account-level rules add a separate safety layer.
Breakeven, trailing stops, partial exits, cooldowns and lifecycle-aware rules increase state and testing requirements.
Sessions, spread, volatility, news or symbol-specific filters affect when otherwise-valid signals are allowed to execute.
One chart and one symbol is structurally simpler than coordinating state and exposure across several instruments.
REST APIs, webhooks, Telegram, dashboards and remote status publishing require additional interfaces and error handling.
Why a specification comes before a quote
The fastest way to get a realistic development estimate is to define exact entry rules, exit rules, risk limits, trade-management behavior, symbols, timeframes and required integrations. Ambiguous rules are not just a pricing problem — they become implementation and testing problems.
A useful way to scope an EA
Instead of asking only “How much is an EA?”, split the project into modules: signal engine, execution, risk, position management, filters, state, diagnostics and integrations. This makes it easier to identify what is essential for version one and what can be added later.
What should be included in delivery
For a custom build, delivery should be judged by the agreed behavior: configurable inputs, defined edge-case handling, logging, error handling and a clear deployment workflow. Backtest results alone do not prove that live operational behavior is robust.