Limitations of the ATAR forecasting model
While there are numerous benefits of using the ATAR forecasting model for new products and marketing, there are also some limitations and cautions that you need to be aware of when using this approach.
It is a highly sensitive forecasting model
Probably the most significant concern with using the ATAR for new product forecasting is that the model is highly sensitive to inputs – primarily because it is a model that tends to multiply the inputs.
For example, if you had an ATAR forecast that was to deliver $1 million per annum in sales revenue using a 10% awareness input – and then if you decided to input 20% instead, then this 10% increase in total awareness would have the impact of doubling the sales revenue forecast for $1 million to $2 million.
Therefore, each time there is a change to one of the input assumptions, they can be a significant change to the bottom line sales forecast.
It relies heavily upon accurate assumptions
As with most financial forecasts, the quality and accuracy of the underlying assumptions are critical to achieving a logical and realistic sales and profit forecast. The ATAR forecasting model is no exception to this rule.
While some of the assumptions and inputs could be relatively accurate (such as price, cost, promotional spend, customer loyalty, and so on), there would be other input assumptions that would be more challenging to produce – particularly for a product quite new to the firm.
For example, if a firm was introducing a brand-new style of product (such as a new category entry), then the percentage of consumers willing to trial the product may be difficult to determine on an accurate basis, even when using a supportive concept test for this data input.
It relies upon multiple assumptions
Compounding the previous limitation of the ATAR in regards to the accuracy of the underlying assumptions, is the scope/number of assumptions needed to successfully forecast sales and profitability using the ATAR method.
While many approaches to financial forecasts will simply use a top level sales number (that is, sales by unit/volume), the ATAR model constructs this top level sales volume by considering multiple factors, such as: the size of the target market, projected awareness, estimated trial percentage, likely availability, repeat/rebuy purchases, and long-term customer loyalty.
Clearly, anytime you have multiple assumptions in place, then there is a likelihood of increased chance of one or more of these assumptions being significantly incorrect – which impacts the overall model and its forecast.
There may be a tendency to “tweak" the profitability or ROI
Marketers may have a conflict of interest when constructing a sales forecast for a new product. This is because, generally (but not always), the marketer is interested in demonstrating that the potential new product has the ability to be financially viable in the marketplace.
This situation – of wanting to demonstrate financial viability – tends to be amplified if the company has already invested in the early development and research of the product concept. Therefore, given the sensitivity of the ATAR model as discussed above, this is a model that needs to be used with care and not manipulated to achieve a bottom line output.
To mitigate the risk of this possibility (of over inflated sales/profit forecasts) thorough and robust justification of the underlying assumptions are necessary.
It locks in the marketing department to key deliverables
Because of the structure of the ATAR forecasting model, it combines financial projections and metrics with marketing inputs – awareness, trial, availability, and repeat/rebuy. Obviously, this is a significant benefit of this approach to new product forecasting.
However, if the previous limitation (of “tweaking" the overall profitability) was a concern/issue, then the marketer who has tried to demonstrate a positive/strong forecast for the potential new product (in order to achieve management approval for launch) may have the challenge of having to deliver potentially unachievable marketing goals.
In other words, if a number of the ATAR inputs were “tweaked" in order to achieve a strong profit projection, and the new product development was approved on that basis, then it is still the marketer’s responsibility to deliver on those forecasts.
This is why, as suggested above, it is necessary to be realistic and robust when setting assumptions and the inputs into the marketing aspects of the ATAR forecasting model.
It may differ from “standard" forecasting approaches within the firm
Because this particular model tends to be designed for new products and for marketing purposes, it is possible that this could be a new/different approach to forecasting in the organization. Therefore, it may be necessary to invest some time in communicating the mechanics of the ATAR forecasting approach and/or the design of the free Excel template on this website (which I would suggest that you should probably use).
It is more accurate for product line extensions and product improvements
Because of the reliance on underlying assumptions to generate a sales and profit forecast, the ATAR forecasting model is very accurate when there is reliable information available. You are far more likely to find reliable information for product line extensions and product improvements, as the firm has engaged in these practices/new products in the past, and these prior market results can be used as benchmarks and assumptions.
Once we start looking at new products that are relatively “unusual" for the firm – that is, they have limited experience and data to draw upon, then we more rely upon generalized assumptions and estimations – potentially leading to a reduction in accuracy, particularly over future years.
Some of the terminology of the ATAR can be misunderstood
While the four components of the ATAR seems simple enough – please refer to the individual articles on each of awareness, trial, availability and repeat/rebuy – care needs be taken when using them.
As a quick example, while availability may refer to the percentage of retailer take up, that may or may not be an accurate percentage to use in the model. If you were selling a simple product like a candy bar, then percentage of take-up in convenience stores and supermarkets would be a good percentage to use in the ATAR calculation (the new product is in 20% of the stores).
However, if you are selling a product that people are willing to shop around for, you need to factor in how many stores that the average consumer will visit. In this case, if you were in 10% of the stores, yet consumers would visit three stores on average, then effectively your rate of availability of the new product could be assumed to be around30% (3 stores X 10%).