Bottom-up Revenue Forecasting. Bottom up forecasting uses company level data to build revenue forecasts. Revenue drivers are industry specific, but will also be dependent on what the company discloses as its key performance indicators. That's KPIs.

For example, if your modeling a Beverages Company, you might forecast units sold and price per unit. For an e-commerce business you might forecast the number of orders processed and average order value. For physical retailer you might forecast the number of stores and revenue per store or square footage of retail space and revenue per square footage of retail space. But in order to do so the company would need to disclose, units sold orders processed and retail stores or retail space. A common issue is that a company might not disclose the actual amounts but may instead disclose the percentage change in these drivers. Here's an example of a disclosure provided by Coca-Cola in its annual report to explain the 2% reduction in revenue in 2013. We can see that most of this reduction came from structural changes that's business disposals and currency fluctuations. However, this is partially offset by a 2% increase in volume sold and a 1% increase in price and mix effects mix effects are a result of companies charging different prices in different regions. That's geographic mix or companies charging different prices for different products or distribution channels. That's product mix. Either way a change in the geographic mix of your revenues or the mix of products sold each year can impact on average selling price. Even if prices haven't actually increased. For example, the 1% increase shown here might be because Coca-Cola has generated a greater proportion of its revenues in developed markets where selling prices are higher or because a greater proportion of product was sold to restaurants and bars rather than supermarkets generating a higher selling price. If we wanted to forecast next year's revenues for Coca-Cola, we can't do this explicitly using the information above as we aren't told what price and volume were last year, but we can do this using total revenue and our expectations for future price and volume growth. To do this, we need to remember that total revenue growth will compound the price and volume increases. If prior year revenue was 100 and prices are expected to increase by 10%, this would give revenues of 110. However, if volumes also increase by 10% then this gives forecast revenues of 121. That's overall revenue growth of 21%. the formula shown on the screen shows how the overall revenue growth is calculated from the two revenue drivers, sometimes instead of forecasting volume and price growth explicitly analysts instead consider how price and volume changes contribute to revenue growth. This is described as the revenue growth contribution. In this situation forecast revenue growth is 21% And there is an equal contribution from price growth and volume growth. We can therefore say that the price growth contribution is half of revenue growth that's 10.5% and similarly the volume growth contribution is 10.5%. But how can we actually forecast price and volume growth or growth contribution? We should start by considering what we know about the business from historic growth rates, but also from the company strategy and management guidance. We can then look at industry dynamics what growth rates are competitors experiencing. Are there any new regulations or trends that will reshape the industry? Finally, we should think about the wider economy our economists forecasting major growth or even contraction in the markets the company operates in and could that affect this company?