One of the key concerns in the software industry is how to deal with the issues of "software licensing". Increasingly customers are putting more pressure on software development firms to provide them with software solutions in a manner that matches their use of the software. This pressure is intensifying as IS managers pay more attention to managing their software costs. Software development firms see the issues from a different perspective. They are greatly concerned with unrealized revenue and excess costs in the form of software piracy, unauthorized use, excess discounts and lengthened sales cycles. nd !oth sides see high administrative costs. These factors have led to a situation where many costly, adversarial relationships have developed !etween software firms and their customers. There is an increasing need for software development firms to !e engaged in the "uestions of "#ow do we remove software licensing as an o!stacle to sales$" and "#ow can we !uild long term, profita!le !usiness relationships with our customers$" These "uestions must !e answered in the face of a rapidly changing technological environment. The answer lies in context and a commitment to an ongoing licensing process as technology and !usiness re"uirements change. Software developers must provide products and services to customers in a manner that fits their customers% !usiness. They must recognize that their customers are not homogenous& that they may have different use patterns of the software and that they may have different !usiness re"uirements. The conse"uences of not taking this approach is to !e vulnera!le to the "price'discount death spiral" and to (eopardize the potential future financial well)!eing of the firm. To uncover the !est solutions to the issues surrounding software licensing, we first have to look at the !usiness transaction !etween the software provider and customers. There are two components of value for which a price is paid* value for function +technical pro!lem solution, and value for product delivery +!usiness pro!lem solution,. -roduct delivery for software is made up of a num!er of components, pieces of which are often referred to as "software licensing." These components are* -hysical delivery -ricing .etrics /iscounts 0icense periods Support and .aintenance 0icense management Tech support 1hange in use level 2ug fixes -latform .igrations -roduct enhancements .etrics +e.g. per user, per node, per transaction, and license periods form the !asis of value while pricing and discounts create the degree of value. Software product delivery is analogous to a manufacturing process where a supplier can rewarded with price premiums for "(ust)in)time" delivery. If the delivery value is inferior or does not meet customer needs, the effect is to devalue the functional or technical value !y giving higher discounts or creating expensive special deals. The apparent driving force !ehind the changes in software metrics is and has !een the pressure applied !y customers on software providers to have costs correspond to actual use. #owever, on closer examination the driving force is more the opposite side of the same coin. 1ustomers have !een primarily reacting to paying for excess capacity. This is caused !y a num!er of things, including situations where metrics and license periods do not match use, purchase for peak use leads to idle assets, and fear of contention for licenses leads to ever purchase. n example of the changes caused !y the pressure from customers has !een the movement from node locked licenses to concurrent use licenses in the 34I5 market. This change however did not satisfy the entire capacity issue. .any customers still have a concern with peak loading and are unwilling to pay for peak loads when their average use is much lower. This has lead some firms to offer monthly use licenses as an option to perpetual or annual use license periods. In addition, some customers have !egun to realize that when purchasing a floating license, they are actually purchasing 67 user hours. This again gives rise to capacity concerns. In some cases, customers want to float licenses across the world for 67 hours or pay less for the licenses they have for use in a specific location. 3nfortunately, the !usiness assumptions +often unarticulated, underlying software providers pricing decisions assume a 8)9: hour use day. In response, many are trying to prevent world wide 67 hour use, again creating an o!stacle to sales. The !ottom line for customers is that, at some level of cost relative to pro!lem importance the unwillingness to pay for excess capacity drives the purchase decision and decreases the software providers% leverage. This, in turn, leads to pro!lems on the provider side. ;hen metrics and license periods do not approximate use, the sales cycle lengthens and !ecomes more costly, and excessive discounting occurs to get the deal. Special deals may also !e created and these often result in high administrative costs. <ven if the metrics and license periods match use, without the software asset management tools to measure the use according to the metrics and periods, a great deal of unauthorized use may occur. This comes primarily from customers not knowing what they are using, rather than outright piracy. s an example, some software providers have sold their software !ased on the num!er of concurrent users without administrative tools. ;ith neither customer nor software provider knowing the actual use, costly, adversarial relationships can develop. The piracy issue also needs to !e addressed since it can !e source of unrealized revenue. 2efore a decision regarding how !est to manage the selected product metrics can !e made, it is important to uncover the software provider%s !eliefs a!out piracy and unauthorized use across account category, geography and channel of distri!ution. It is important to articulate !eliefs regarding each of the intersections to formulate appropriate !usiness decisions. =or example, many software providers !elieve that piracy in not a ma(or concern within ma(or accounts in the 3.S., even though unauthorized use does occur. In this situation, !enign license management may !e much more effective than the hard core control and compliance re"uired in the Third ;orld to com!at piracy. The !ottom line for software development firms is that at some level of revenue loss or excess costs, the unwillingness to suffer this loss drives the adoption of new delivery methods and appropriate license management. Software metrics and license periods that approximate use along with appropriate license management decisions can !ring the !usiness transaction e"uation !ack into e"uili!rium. ;hen this happens the !usiness transaction gets less costly and more profita!le for !oth sides. The metrics or !asis of value decision is the starting off point for esta!lishing value for product delivery. The initial step is to determine whether a software product should !e considered as a "tool" or as an "asset" that should !e managed according to a dynamic usage metric. Software products that that fit the "tool" category are usually relatively low cost and'or the product%s value is driven !y the a!ility to solve potential pro!lems at any time rather than actual use. Typical products that fit this category are word processors, spread sheets and even compilers. Typical metrics are either node !ased or named user +not floating, !ased. =or these static metrics, the num!er of licenses re"uired is relatively easy to determine for the salesperson and the customer. In most cases these metrics are much easier to administer than dynamic licenses. s value increases, products are viewed more as assets. In order to manage these assets well, dynamic licensing metrics need to !e chosen that closely approximate use or "work accomplished". These software products can !e segmented into categories according to functional characteristics and usage characteristics& different licensing metrics may fit each of these segments. 1lient)side software tends to !e interactive intensive while server)side software tends to !e more compute intensive. =or interactive intensive applications the performance of the system running the application has relatively little effect on the use or amount of work accomplished over a specific period of time. =or these applications, license metrics +e.g. concurrent users, should !e independent of processor performance. 1onversely, for compute intensive applications that operate primarily on a server, processor performance significantly effects the amount of work accomplished over a specific period of time. =or these applications processor performance must !e taken into consideration when selecting metrics if revenue to the software developer is to correlate closely to work accomplished using the software. Some developers of this application category are selling their products on a user !asis, !ut the cost per user varies with system performance. /ynamic licenses can re"uire a much greater level of administrative support and can !e more difficult to sell. ;ithout good usage data, neither the salesperson nor the customer may !e a!le to determine the num!er of licenses re"uired to support the users on a system at the initial sale. s a result, the sales cycle can !ecome longer and more costly. To effectively sell products according to dynamic metrics generally re"uires a dedicated sales force that has a good understanding of customer usage patterns. 2ecause of the difficulty in selling dynamic licenses and the potential administrative !urden, it is sometimes wise to consider static metrics that approximate the underlying use. One example of these static metrics is from the human resource market where the software is sold !y num!er of employees. This metric approximates the underlying work accomplished regardless of where the software is running or how many users are accessing it at any time. The metric is easy to understand, sell and administer. The key to metric selection is to choose metrics that approximate use where use is defined as work accomplished. 1urrent literature emphasizes the !enefits of user)!ased licensing, !ut the decision maker must !e aware that users approximate use in only certain situations. =or instance, user !ased license metrics may approximate use well for O0T- applications, !ut when the same product is used at night in !atch mode the approximation no longer holds up. .etrics, license periods and metric management decisions can all !e made independent of pricing decisions. #owever, this is the critical point where the process of change to new software delivery methods often !reaks down. The key element that is missing is customer usage data. ;ithout via!le usage data, it is difficult for the software provider to select a price for the new metrics. ;ithout usage data, the 1=O at the provider can, and often will, stop the process until the impact to the !ottom line can !e determined. Often the customer also is not aware of usage patterns for specific products and as a result has a difficult time determining if the new model and the price is a good deal or a !ad deal. This puts another road!lock in the path of a successful !usiness transaction. The customer is in control of the usage data, !ut may or may not have access to it. This again speaks to the need for software providers to develop strong long)term !usiness relationships with their customers so that usage data can !e shared. ;ithout this sharing of usage data, the road!locks and adversarial relationships will continue to !lock successful !usiness partnerships. Software license management systems, originally developed as tools to ensure control and compliance, now are transforming into software asset management systems. They can !e used to collect usage data and ensure revenue integrity for the software provider. -rior to making any shifts in licensing methods, it makes sense to use software license management systems to gather usage data so sound !usiness and financial decisions can !e made later as the new metrics are introduced in the future. The Winning Formula Software providers must !e committed to providing solutions to their customers% pro!lems through advanced technology and new delivery methods. >ust as corporations make an investment in technology, they must also !e committed to making an investment in delivery methods. =or success, software providers must recognize they are engaged in an ongoing process that must !alance their needs and their customers% needs with advances in technology. nd lastly, !ut perhaps most importantly, software providers must !e committed to providing products and services to customers in a manner that fits their customers% !usinesses. ;ith the context of this formula in mind, software providers can remove software licensing as an o!stacle to sales and can focus their efforts on developing long)term, profita!le !usiness relationships with their customers. It is these relationships that results in handshakes rather than adversarial situations, which in turn result in success for customers and software providers.