One of the biggest pitfalls when developing a forecast is “spurious accuracy”. Teams (and investors) become obsessed on building plans based on very detailed granular assumptions. For example an individual POS will sell “X” or “X+1” per week or indeed a sales person will secure 1 or 2 contracts. Individually each granular assumption is logical but can be argued up or down by small amounts. This process creates a granular plan that can only be precisely wrong as small decisions on a POS or sales person level are multiplied up. In reality a good planner avoids spurious accuracy and knows that any plan is likely to have compensating errors.
Good plans are build up from the bottom up and the top down. With the result sense tested. Early plans are better to be reviewed at the aggregated level where the broad impact of the planning decisions can be assessed and adjusted. Try and aviod getting bogged down (not seeing the wood for the trees) and take the "helicopter" view.
A good plan should be the most likely scenario with alternatives being covered by sensitivities or scenarios.