On May 25th, Tensoft hosted a webcast* with Silicon Valley software revenue recognition expert Jeffrey Werner entitled: Revenue Recognition Accounting for Software as a Service (SaaS). This is the second of five blog entries detailing Werner’s responses to audience questions posed after the live webcast.
Question - Best Estimated Selling Price
How are you seeing the “Best Estimate Selling Price” (BESP) working in practice, especially where there’s no pricing history and management doesn’t establish firm pricing methodologies? Should we just bundle everything together (set-up, monthly service, etc.) and recognize over the longer of contract or expected customer life?
Response
BESP is required. If there are multiple elements that are separable, then the company must come up with estimated selling prices using the best available information and analysis.
The elements have to meet the separation criteria in order to have BESP applied. SaaS arrangements often do not have separable elements because the elements do not have stand-alone value. An example is setup fees.
There seem to be a few approaches that are being used consistently for elements that can be separated.
In companies with significant hardware costs and low volume, high-dollar products are often approaching BESP using a gross margin approach. This approach takes the cost of each element of a transaction and adds the expected or average gross margin to arrive at the estimated selling price. The gross margin is often at a division or product family level. Some companies with fewer product offerings might use a company-wide gross margin.
Other companies are using a discount from list price approach. This works when the company has a consistent pricing approach and a discount range by product or product family that has some consistency. Companies look at the average discount and then apply that to list price for each product to arrive at the BESP.
Some companies are using what I call a "VSOE Light." PwC calls this the “broken” or “failed” VSOE approach. These companies apply the same analysis as a VSOE study but allow greater variances in the covered population and the plus or minus percentage. For example, if the pricing of a product is consistent for say 60% of the population with a 20% plus or minus variance, that might be a good indicator of a Best Estimated Selling Price. This would compare to the normal VSOE analysis of 80-85% of the population covered with a 10-15% variance.
Obviously, there are complexities to these approaches and often there are “devils in the details.” Companies should work with consultants and their auditors to develop a reasonable approach.
* Click here to view this on-demand webcast in its entirety.
