I received a recent referral from an executive at Microsoft to check out an innovative, cloud-based forecasting as a service provider, and I’m glad I did.  Manufacturers, retailers or service providers with complex forecasting needs may want to check out Lokad.  After a briefing with CEO Joannès Vermorel, I came away with an impression that mid-market and even larger firms have an interesting and somewhat cost affordable alternative in generating more timely and accurate forecasting.

Lokad, a Paris based firm, has been in existence for two years, and from its inception, has provided a cloud-based service offering which is somewhat unique in forecasting needs.  To date, this vendor has garnered 300 customers, the majority of which reside in Noth America.

Customers are provided alternatives in loading demand data over the web. CEO Vermorel noted that the company differentiates itself on the sophistication of its staff of mathematicians who take on challenges reflected in difficult forecasting problems.  Customers are asked to load data “as it is”, avoiding the need to perform tedious data formatting and pre-analysis.  Lokad believes in a different approach than many existing ERP or best-of-breed forecasting applications, one that considers other products as well as the target product.  An example noted on its web site notes an example of forecasting the sales of chocolate bars, which might include the analysis pegging seasonality patterns of other similar chocolate bars. Once forecasts are calculated, the Lokad technology delivers forecasts via an open Applications Programming Interface (API) to SOAP compliant applications, which can be many applications such as demand planning, vendor managed inventory, replenishment applications or inventory optimization.

As a Microsoft Azure based cloud based provider, Lokad can leverage large numbers of blade servers and parallelization computing routines to crunch high amounts of data in a far more rapid rate than many in-house applications running on fixed IT applications server. The ability to call on extra computing and analysis power is rather flexible and scalable, especially if your forecasting or planning processes requires rapid turnaround.

The other compelling option for consideration of Lokad is its cost. Pricing is dependent on the number of forecasts a customer consumes monthly and ranges from $70 per month for 10,000 forecast equivalents, or $1500 per month for one million forecast equivalents.  We believe these costs can be fairly reasonable, since a small or mid-market company with typical forecasting requirements can obtain this service in a range of $100-$200 per month. Large companies may also find Lokad to be a more cost affordable alternative when process needs call for more timely and sophisticated forecasting.

Bob Ferrari

Disclosure: Neither Lokad or Microsoft are current clients of the Supply Chain Matters blog.