Pricing for Software Product Managers
Director, Product Architecture
SanDisk, Secure Content Solutions
Table of Contents
- Pricing Software Products
- Discounting and Non-Linear Pricing
- Pricing Discrimination
- Rebates and Temporary Discounts
- Other Pricing Issues
- Pricing Mistakes
- Appendices and Resources
The Need for Pricing Tactics
Pricing has far reaching effects beyond the cost of the product. Pricing is just as much a positioning statement as a definition of the cost to buy. Pricing defines the entry threshold: who your buyers are and their sensitivities, which competitors you will encounter, who you will be negotiating when selling the product and what the customers expectations will be.
Pricing is unique from other marketing decisions for several reasons:
- Price is the only marketing element that produces revenue. All other marketing decisions produce costs.
- Pricing is the most flexible marketing decision.
- Pricing reflects a product’s strengths and weaknesses. It implies value and positioning.
Good pricing tactics will remove the price issue from being an obstacle to a sale.
When pricing software, using the Economics 101 that is taught in college to set a products price is irrelevant. There are many reasons for this:
- Estimating price elasticity for specific product is practically impossible. Hence basing pricing decisions on supply and demand curves is impossible.
- Estimating the potential market for a product is possible but estimating demand is problematic. Most customers tend to be enthusiastic when seeing a new product. Their input is not a good indicator for real demand.
- For enterprise software, sales numbers are too small for a statistical significant study. By the time a company has sold enough licenses, it has advanced on to a newer version or the market has changed.
- For most products, there are typically many viable competing products and their influence on the demand curve is hard to estimate.
- Product life cycles are too short, making comparisons more difficult.
- Purchase decisions are complex and are influenced my many, constantly changing factors.
When setting the price for a software product, classical economic theory comes up short. Here is an empirical, iterative method to come up with a price.
Guidelines for Setting the Price of Software in Existing Markets
The purpose of these guidelines is to arrive at the “right” price. This is the price that lets the company accomplish its goals for revenue, profit, market share, renewals, etc. The method detailed below will help you identify the highest price a market with existing competitor presence will bear:
- The price of the software must be less than the ROI it provides. The smaller the ratio of the ROI to the cost, the easier the sale.
- Create a market segmentation chart based on feature sets. Identify all competing products and place them on this chart. Identify and group the value elements in the product that address the needs of each of segment.
- For each segment, identify the features that customers are willing to pay extra for, that differentiate your product from the competitors. Attach a price tag to the value of each attribute that is not identical such as:
- The feature and functional differences.
- The difference in brand value that customers attribute to the products.
- The difference of cost for implementing the respective products.
- Any other item that customers attach value to such as localization of the application, geographic proximity (for services) etc.
- If the product excels in a certain aspect, than simply add that value to the price, if it lags, simply subtract the value. This step must be iterated for each competing product. The price of the software must be similar or less than that of the main competing product in each segment minus the difference in price that are justified by the functional and other aspects previously identified.
- The price must be below the purchasing authority of the targeted decision maker signing off on the purchase.
- The price must be outside the Death Zone of $5000 – $20,000. For details on the Death Zone, see below.
- The price must fit how the market perceives the product category. For example, desktop utilities up to $50, productivity tools, up to $500 etc. If the product is priced too high, the price will become an issue. If it is priced too low, customers will perceive it as not worthy.
Guidelines for Setting the Price of Software in New Markets
If there are no reference products, the approach is slightly different. The first step in setting a price is identifying how customers will position it in their mind. If the product is seen by customers as a utility or productivity tool, price in these ranges. That is, until you the product can be positioned (in the buyers minds!) as belonging to a higher place in the food chain. See below for examples of products and typical price ranges.
If the product does not fall into the previous category, start by defining the price ceiling. This is the highest price based on a products benefits. A high price will work if early adopters are willing to pay a premium for a new product. However this price level may prove to be unrealistic as there may not be a sufficient number of buyers for a new product at that price level.
Then, choose a penetration price. Penetration pricing is used when a product is first launched in order to gain market share. A low penetration price is used to discourage competitors from entering the market and to gain market share. Its drawbacks are lower margins, difficulty in raising prices in the future because long term pricing expectations are now set and the risk of customers perceiving the low price as a low quality indicator. The penetration price has to be sustainable and higher than the company’s variable costs. If possible, the price should be low enough as to remove the price of the product from the buying decision.
These two markers set the price range for a new product. Follow the relevant guidelines in the previous section to finalize the price point.
Comments on Setting Prices
- Before making pricing decisions, you must thoroughly understand your target markets decision making and buying processes.
- Properly priced software will not guarantee the company’s profitability.
- The price has to take into consideration what the customers feel is reasonable. For example, market leaders are expected to charge more hence their higher prices can be perceived as legitimate (up to a certain level).
- When deciding on which product competes with your own, the markets perspective is what counts.
- Internal company parameters such as distribution costs come into play only when looking at the potential profitability of the product. I.e. can the company make money selling the product at the price point that was found in the previous process If it does not fill the following criteria, the process above must be repeated.
- The price of the software must be higher than the cost of selling it and the margins must be high
er than the cost of creating, marketing, selling and supporting it. Or else the product will lose money.
- Using the pricing model as a differentiator is always worth considering as long as the model is easy to explain and it makes sense to the customer.
- The cost of training, implementing and supporting the product are perceived as additional costs by the customer.
- When new to a market, being a small, unknown company minimizes brand value. Lower, penetration pricing may be required.
- If customer segments value the product significantly differently, this may justify segmenting the product for each of these markets.
- When attempting to price commodity products, it is basically, the competition that sets the price of the product. Setting a higher price in a commodity market is limited to the company’s ability to differentiate the product from its competition. On the flip side, offering a lower price in such a market is sustainable only if the company has a lower cost structure.
- When setting the price within a range of competing products, it is important to understand how reference prices affect your customers price evaluation. This is imperative when customers have limited product or price knowledge.
- Just like other aspects of the product, pricing needs to follow the technology adoption life cycle that the product follows. Early market buyers may be interested in your product but tend not to be willing to pay its full price. It makes sense to price the product at its target price for larger markets and offer early adaptors the product at a discount.
The Software Price Dead Zone
The Pricing Dead Zone is a price range between $5,000 and $20,000. Some would even say that the range extends up to the $40,000 – $50,000 range. Software products in the Dead Zone are an exception. This is because software priced in this range is hard to sell profitably. Products that cost less than ~$5,000 can be sold over the web or through channels. A purchase of this size is within the decision authority of middle managers and there is no need for on site visits to close a sale. More expensive products require higher level signing authority or purchasing committees. A committees decision can cause the sales cycle to drag on for months and get entangled in internal politics. These products require sales reps on-site visits but have to produce enough profit to support this type of sales effort. $20,000 is at the bottom of the price range that can support a complex sales process. The exact extent of the Dead Zone depends on the specifics of how the product is sold.
Typical Software Price Ranges
Utilities cost up to the $50 – $60 range. Purchases of utilities are many times spur-of-the moment decisions. Customers need to feel comfortable that possible financial risk of buying the wrong product is minimal.
Productivity tools $100 to $500. These are purchases that within the budget of a low level manager.
Professional tools $1000 to $5000. These are applications that are required by professionals to do their job. These applications include computer aided design tools such and many others.
Enterprise applications – $20,000 and up. These are applications that impact many functions and departments in the customers organization and that require an evaluation process and sometimes, a purchasing committee. Selling into such a customer is many times a costly and labor intensive process.
Perpetual vs. Subscription Licensing
Subscription software is an application that is rented on a temporary basis. Licensing is usually on an annual basis but monthly terms are available as well. Salesforce.com is a classic example of a hosted product that is priced per user per month. Subscription licensing works when customers see an ongoing benefit from the software. From the customers perspective, it lets them buy into the product while minimizing their initial investment and exposure. From the ISVs perspective, it keeps them focused on making the customer successful with the product rather then the “fire and forget” approach to selling software.
Moving from a perpetual license to a subscription model increases the vendors risk as it becomes easier for customers to bail out on them. It may also have a negative effect on the short term stock price due to Wall Street’s focus on quarterly revenue vs. cash flow as the vendor is “mortgaging their present for their future.” This is because, over time, the income statement reflects the growth from prior years bookings, as the deferred subscription revenue is transferred to the income statement. Over the long term, the subscription model allows for significantly better revenue visibility and consistency. This is beneficial as Wall Street loves companies that make their numbers. For example, when signing a three-year license for $100,000, one twelfth of it can be recognized each quarter with high certainty. In such a case, cash flow becomes the much more representative indicator of income. This works as long as the renewal rate is high.
By offering a subscription pricing model, customers face smaller payments. From the sales reps perspective, a lower initial price lets them aim their pitch lower in the customers organization. Another advantage for the sales process is that a calculating an ROI on a shorter time scale makes it more tangible, hence helping the sale along.
One implication of offering a subscription model is that the vendor is betting their future on their ability to keep customers. For hosted apps, setting up a hosting environment, can be very costly and by doing so, the vendor is basically giving your customers a loan that will be paid back over the length of the contract. This creates additional risks that vendors may want to steer away from.
So now, to the bottom line. How much do you charge for a subscription model There are no axioms here but many companies charge one third of the cost of a perpetual license for an annual term. When offering a subscription model, maintenance is usually mandatory.
Pricing Maintenance and Support
For enterprise applications, 18% to 20% of the list price is the standard cost of support . This usually includes support over the phone for a single contact from the customer during regular business hours as well as releases (both point and major releases). More advanced packages that include 24/7 support are priced higher, in the 20-25% range and require a minimum of $30,000-$100,000. Minimums of $200,000-$300,000 are the norm for packages that include assigned support engineers. Onsite support should always be priced as an extra.
Most companies have a no-discount policy on support. That is, even when the software is discounted, the support pricing stays at the standard percentage of the list price. Very large deals may justify a discount, for example, if all support calls are routed from a single person at the customer. One approach is to give away a few months of the first year. Psychologically, its better to give away months than to lower the price of the yearly contract.
For non-corporate users, there are two basic models for providing support:
- The per-incident model: The most common model for personal support is per incident that is, a flat rate for resolving each support question, regardless of call length. The median per-incident price for surveyed companies  that offer this option is now $100, with 50% of these companies charging per-incident prices between $35 and $185. Support for developer tools and more technically advanced issues run into the hundreds of Dollars per incident. For example, a call into Microsoft’s tech support for developers costs $245. (From experience, they provide great service!) These models typically include a refund if the problem is determined to be a defect in the vendors product.
- The per-minute model: A less-popular model is a per minute rate. Here, there is less variation in pricing: The median per-minute price is $2.71 (18 respondents) and the 50% range is $2.00 to $2.95. Note that the $3.00 per-minute rate is one of the few service prices where there’s significant customer sensitivity and push back.
Discounts come in two variations, scheduled and negotiated. Scheduled discounts are those that are pre-approved by the company, based upon pre-defined criteria such as the volume of the purchase. Negotiated discounts are an ad-hoc result of the sales process that differ from or go beyond the pre-set scheduled discounts. This article will only cover scheduled discounts.
There are multiple reasons why ISVs offer volume discounts:
- Many times, the utility to the customer of additional licenses decreases as volume increases. To guarantee that the value to the customer is more than the price of the software, the price must decrease as the volume goes up.
- In most sales situations, the cost per unit of selling decreases. This savings can be then passed on to the customer.
- A volume purchase increases the customers investment in your product and reduces the chance of their buying the competitors product.
- Large customers are convinced that it is their God given right to pay less per unit than smaller customers.
- Buying more units now than in the future has a discounted current value.
Once a discount is offered, buyers will assume that discountor a better onewill be offered for all future purchases.
Calculating Volume Discounts
The way most companies calculate their discount schedules is surprisingly off the cuff. They simply decide how much money they would like to get from a large target customer per user and then draw a curve between the price of one unit and the price of a unit at the high volume level. They then stand back, look at the curve and play with it until it looks good.
Another, more rigorous method for calculating volume discounts, is to select a consistent discount rate for every growth in units. For example, a 10% discount on the 10 20 units, a 10% discount from the previous price on the next 10 units (=a 19% discount from the original price) and so on.
For three slightly different ways of calculating the price of a volume discount between price tiers, see Appendix B.
MarketShares report on discounting in the software industry  found that:
- Discounting is widespread, and significant. All the companies surveyed reported using discounts, and those discounts averaged nearly 40% of list price.
- Company growth rates appear to be inversely related to the extent of discounts allowed by respondent companies, with respondents in faster growing companies reporting discounts that were nearly 20 percentage points less than discounts given by slower growth companies.
- Nearly 75% of respondents acknowledged one or more important benefits to monitoring discount activity.
- The actual extent of discounts given is not generally known nor documented by respondents. Less than 50% of respondents reported that they track discounts either frequently or regularly.
- Negotiated discounts were significantly lower in companies where discounting is tracked. Negotiated discounts by “trackers” averaged approximately 7 percentage points below what was allowed by “non-trackers.”
Value Added Resellers (VARs) get the software they resell at a discount. Discounts are typically between 40% and 60%, depending on the marketing and sales efforts required by the VAR to promote the software. Many companies incentivize VARs by creating volume thresholds that increase their discount level. Tier discounts require VARS to commit to sales volume. For example, a15% discount for no commitment, 35% for very serious commitments. VARs receive training & licenses for an additional cost. Just as a comparison, reference partners whose activities are limited to referring customers to the vendor, get 5-10% of the deal.
Royalties paid to OEM a product run anywhere from 10% to 40%. Some companies require an up front fee for Non Refundable Engineering (NRE). NRE are efforts needed to tailor the product to the OEMs specific needs. NRE fees include charges for developers, QA and project management. These fees can easily run into the six digits. Some OEM deals will tier their pricing based on the size of the up-front fees and volume commitment. As a rule of thumb, the higher the commitment and up-front fee, the lower the royalties.
Site licenses are a model that gives customers unlimited use of a product across their enterprise while paying a flat fee. A buyers request for a site licenses it mostly a purchasing ploy. Their reasoning is that with a site licenses they dont have to worry about counting seats. However, its only another way to ask: what is your best price One problem with this model is that as a vendor, you lose your ability to track the number of installations at the customer site, and if your product is successful, you will be leaving money on the table. Another drawback of site licensing is that when you sell a site license, you have effectively lost that customer for any repeat sales. If you are concerned about getting the product in front as many users as possible, just offer steeper discounts to encourage proliferation and use. Hence, rule #1 for site licensees for vendors: avoid them.
From a pricing perspective, products that are sold to academia can be divided into two. Products that are used for teaching and that a company can expect the students to purchase later in their professional career and all other products. For the former, companies tend to price their products at a deep discount. For the latter, ISVs usually offer up to 40% discounts. A slightly different approach is for ISVs to start academic pricing at the second copy price. In other words, the discount equals what they would give normal commercial users for purchasing a second copy.
Re. hardware products due to the lower margins, vendors cannot offer the same level of discounts that software vendors offer. Whatever discounts are available, they are much smaller.
Price discrimination is a technique for maximizing profits by offering the same or similar product at different prices to different customers. Price discrimination can be explicit or implicit. Explicit price discrimination is when a special price is limited to customers who meet certain criteria. For example, academic pricing is a form of explicit price discrimination because only students and faculty can buy at that price (at least in theory). Implicit price discrimination is when all customers are technically eligible for the special price, but the vendor inserts a condition that makes it unattractive to some. For example, rebate programs are a form of implicit price discrimination.
The justification behind price discrimination is that different market segments value the product differently and will therefore be willing to pay varying prices. If segment A values the product at $1000 and segment B values it at $500, when the price is $900, only segment A will purchase it. If the product is priced at $400, both segments will purchase but with segment A, money will be left on the table.
To make Pricing discrimination work:
- Each segment needs to have a specific version. This can be done by creating specific functionality for that segment.
- One market segment cannot buy the product created for another seg
- The difference in pricing must be justifiable and must not create a feeling with customers that they are being treated unfairly.
Academic pricing is one example of price discrimination where the same product is sold at a discount not available to other market segments. This works because other segments hold a common belief that education is important and that businesses are expected to support it. International pricing is another example of pricing discrimination.
Another common form of pricing discrimination is introductory pricing. The idea behind this technique is to release a new product at a price premium and to lower the price in time. This is a common technique in the computer chip industry where power hungry buyers are willing to pay a premium for the latest and greatest in chip technology. The reverse can also be true: introduce a product at a significant discount for a limited period to stimulate early sales and then return to the higher list price once the initial surge of excitement has passed. Introductory pricing is generally limited to one or two quarters.
Illegal Pricing Discrimination
The Robinson-Patman Act made it illegal for sellers to directly or indirectly discriminate in the price of similar commodities, if the effect hurts competition. This is especially important when selling to distributors and VARs. For example, if a vendor has two distributors that compete with each other, they have to offered the same basic terms. If one distributor is allowed to buy software from you at a lower price than the other, competition is adversely effected because the second distributor, buying at the higher price, will have a greater difficulty in reselling the software.
A detailed discussion of the implications of the Robinson-Patman Act is beyond the scope of this text but it is important to note that there are situations where pricing discrimination is explicitly legal. These include situations where the vendors manufacturing, delivery or financing costs are different for different customers as well as situations where a competitor dropped their prices. Meeting the lower price is not illegal even if this price is not offered to other customers.
Note that the law applies only to products and not to services.
Prices of the same product in international markets many times vary compared to US pricing. The uplift (as its called) varies anywhere from zero to a premium of 50%. This uplift is justified by increased costs due to the need for localization of the product as well as marketing and sales expenditures the vendor faces in foreign markets. The cost of localizing the software has to be considered but in many cases is not the bulk of the investment in foreign sales. Higher support costs are due to the additional languages needed, the more expensive labor (at least in Western Europe) and of course increased business risk. On the flip side, in some geographies such as in Asia, services are less expensive than in the US.
Differential pricing in international markets runs the risk of creating a gray market for the product.
Another issue that makes international pricing difficult to manage is the fluctuations in exchange rates. There are two approaches to adjusting prices when the exchange rates change:
- Adjust the local price to reflect the price in U.S. Dollars. This approach may cause difficulties in countries where the currencys buying power decreases compared to that of the Dollar.
- Adjust the local price to partially compensate for the change in the exchange rate.
Both should be done with an eye on optimizing sales, taking into consideration how revenue is affected as well as the effect the change has on gray market pressures.
Bundling is when a group of products (or services) is offered as a single package. By offering bundles ISVs can increase their sales to segments that would buy only one product. There are two types of bundling:
- Product bundling. Product bundling is when two products are integrated into a single package. The purpose f product bundling is to create a combined product that has more value to customers than the separate parts. An example of product bundling is the Oracle ERP package where the database and application are bundled into a single package.
- Price bundling. This is when an ISV provides a discount to customers that buy two or more products at the same time.
There are basic differences between price and product bundling. Whereas price bundling is a pricing and promotional tool, product bundling is more strategic in that it creates added value. Price bundling products does not create added value in itself. Therefore, a discount on their combined prices has be offered to motivate consumers to buy the bundle.
Bundling can be pure or mixed. Pure bundling is when a vendor does not offer any other option but the bundled products. Pure price bundling is basically forcing a customer to buy at least one product that they are not interested in and can be illegal. See below for details. Mixed bundling is when a vendor offers both the bundle and the products separately.
Price bundling is used to:
- Increase sales to segments have different perceived values for the vendors products.
- Expose a new product to a large customer base.
- Provide product visibility and a low cost opportunity for customers to test a new product.
- By offering bundles, vendors make it difficult for consumers to price-shop.
Product bundling is used to create added value for customers. By using integrated products, customers can increase productivity, performance, lower costs of ownership and reduce purchasing costs.
In both types of bundling, vendors can increase customers switchover costs by selling them more products than they would have bought originally.
An Example Price Bundling
Lets look at two Microsoft products Word and Excel that cost $250 each when purchased separately. Assume a market segment of office managers that value Word at $350 and Excel at $100. Because Word is priced lower than its perceived value, this segment will buy Word. On the other hand, because Excel is priced higher than its perceived value, this segment will not buy Excel. Assume a second segment of accountants. By contrast, accountants may value Word at $150 and Excel at $300 and will thus buy Excel but not Word. Also, for the sake of simplicity, well assume that each segment has one member. In this scenario, each segment will buy one of the products, resulting in $500 total revenue.
A more profitable scenario can be created by bundling both products for $400, a price point that places both products beneath each segments value for both applications. Both segments value the bundle at or above the bundle price at $450, and therefore purchase the bundle. With this approach, the total revenue for Microsoft is $800.
There are several requirements for successfully implementing a pricing bundling scheme:
- The products are complementary and not substitutes to each other.
- Individual segments have different perceived values for the specific products but similar overall perceived value for the bundle.
- The unit costs for the parts of the bundle must be sufficiently low so that selling bundles at a discount is more profitable than individual products.
- There is no coercion of customers to buy something they do not want.
Compensating for Bundling affects on Profits
An issue that has to be addressed when offering products as a bundle, is the potential lose of revenue. The higher the variable costs for the products in a price bundle, the larger the increase in sales needed to overcome the discount involved. For example, consider two offerings: a refrigerator and stove costing $2000 and $1000 with variable costs of $1600 and $800 respectively. The second offering is a bundle of a spread sheet and a package of financial macros costing $300 and $100 with variable costs of $20 and $10 respectively. When sold separately the packages will provide $600 and $370 of profit. By offering a discount of 10% on the bundles, the profit will be reduced to $300 and $330 respectively. Therefore, the appliances company has to sell 100% more units to make up for the discount vs. the software company that has to increase sales only by 9% to make up for the discount.
Legal Issues When Bundling
All mixed bundling strategies are legal. This is because the customers ability to choose the product they want is not hindered.
On the other hand, pure pricing bundling is illegal if the vendor has market power. Market power means that the vendor can force a consumer to do something that he would not do in a competitive market or when a substantial amount of commerce is at stake.
If a vendor possesses market power, pure product bundling is legal only if the benefits to consumers offset potential damage to competition. For example, in the Microsoft vs. DOJ, Microsofts claim of consumer benefit was enough to justify the integration of Internet Explorer with Windows. Note that merely combining products together in a single installation does not constitute integration and will be difficult to defend as providing benefit to consumers.
Unbundling is a process where a product offering is split up into modules with some modules becoming optional or by adding optional modules. By taking a complex product and splitting it into modules, the product can become attractive to additional segments. Furthermore, each segment tends to become less price sensitive regarding the modules they need. For unbundling to make business sense, sales to additional customers have to make up for the optional modules that customers passed over. Another drawback of unbundling is the added complexity to the product. After unbundling the product, there are multiple options for customer installations, managing the product and supporting it increase in complexity. One risk of unbundling is that if the product becomes too granular, vendors run the risk of giving the customer a feeling that they are nickel and diming.
Rebates are used to stimulate short term increases in sales and for enticing price sensitive buyers that would otherwise be reluctant to buy at the regular price. A temporary discount can increase customer demand for a product. However, this peak in demand is usually temporary and will many times decrease future short term demand. Price promotions may entice new, price-conscientious buyers but they can actually hurt future sales to the existing customer base. Promotions are tactical, not strategic, and they need to be managed that way.
By reducing the price of a product, ISVs reduce the risk to consumers trying an unfamiliar product. Assuming that the consumer has a good experience with the product, they will be more likely to purchase it the next time, even without a rebate. This is especially true if by using the product, there are significant switching costs for the customer. From a competition perspective, rebates and other forms of temporary discounts are used to lower prices while attempting to avoid a price war.
When executing a promotion, vendors have to beware of the following:
- By reducing the price of their product, even temporarily, vendors risk implying that their product has inferior value.
- If a temporary price promotion goes on for too long customers may begin to expect the lower price. The reference price is then perceived as expensive and customers are reluctant to pay it.
- The promotion must be targeted to new buyers and not to repeat buyers.
One way to meet these criteria, is by creating a trial offer. This is basically a type of temporary discount. Usually, a condition is attached to emphasize the trial offers special nature. This can be done by setting an expiration date to the offer, requiring an additional purchase (a form of bundling) or an exchange of something of value. For example, the customers agreement to present at a trade show on the vendors behalf.
Natural price points are prices at which there are discontinuities in the price / demand curve. Customers expect to see commodity software products priced at these natural price points that are traditionally, $19.95, $29.95, $49.95, $99, $199, $495, etc. The effect of increased demand for $19.95 mouse vs. one that costs $20 may stem from an underestimation mechanism. One explanation is consumers’ tendency to round prices down and to compare prices from left to right. 
As a side note, the $.95 of the price tag was created to force cashiers to give customers change.
A price war occurs when two companies, drop their prices regularly to close sales. Pricing wars start once the differentiation between products has eroded. Proper product planning and positioning can help prevent a price war by allowing the leading vendor charge a premium. However, if two competing products have similar offerings, as the market matures, price becomes a bigger factor in the buying decision. Unless the vendors can extract themselves from the price war with better positioning, the vendor that is able to offer lower prices over time will win the war.
Reducing prices dont necessarily cause a pricing war. This depends on where the product is positioned. If it is the highest priced product, dropping the price to be more competitive will probably not result in a price war. If you are the lowest priced competitor, you may serve a different customer segment and competitors may not respond.
Some factors that increase the likelihood of a price war in a given market are:
- A perception by managers that pricing is an easy to implement and reversible tactic.
- Commoditized products that customers cannot differentiate between them.
- Multiple competitors and manufacturing over-capacity.
- Low switching costs between products.
- High price sensitivity.
One of the ironies of price wars is that while price wars are usually started as an attempt to increase market share, when the dust settles, the respective market shares of the players tend to remain constant but at a lower prices and margins.
A gray market is created when a product that costs less in one market is sold to another market where prices are higher through unofficial channels. There are multiple pressures that can foster the creation of gray markets. These include price differentials and an authorized retailer can’t sell all its inventory may move the leftovers to unauthorized dealers.
A classic example is the current sale of pharmaceuticals from Canada to the US. Canada places limits on the prices of pharmaceuticals and so they are significantly cheaper than in the US. This creates a strong motivation for US customers to import drugs from Canada. Another example is Amazon. Books sell for different prices on the local Amazon websites (see:http://www.pricenoia.com/ for a comparison). What prevents a German customer from buying a book on the American Amazon website and having it shipped to their home in Germany if they can save money When it comes to consumer software, where delivery costs are zero, the problem may is more pronounced. At the time of writing this article, Norton Anti Virus costs $49.95 in the online store for US customers, $82 (44.99) for British customers and $76.42 for the Norwegian version. As most customers with a basic knowledge of English can get away with the cheaper US version, Symantec runs the risk of creating a gray market for their products. In this case, European customers buying the US version of the software.
Gray markets can damage channel relationships but the aspect that is relevant to this article is the undermining of the segmented pricing schemes. A fundamental aspect of international pricing is the ability to price at the level that each market can bear. When a gray market forms, it can limit the company’s ability to charge a premium in the given market.
In some situations , gray market sales outstrip authorized sales. In Malaysia, cell phones purchased on the gray market account for 70% of total cell phone sales. Similarly, in India, sales of gray market personal computers outnumber authorized sales by two to one.
Gray markets aren’t always bad . As long as they do not directly clash with the existing channels, there can be beneficial sides to them such as incremental sales and the ability to reach into untapped markets.
How Pricing Affects Sales Methods
As a rule, the lower the price of the software, the less effort vendors can spend selling it. Enterprise applications are costly to sell. Marketing programs for generating leads can cost anywhere from $20 to thousands of dollars per lead at targeted trade shows. Then there is the time spent by the sales rep qualifying the prospect and traveling to meet him. An onsite sales call typically costs $2000-5000. It costs roughly $2000 per day to send a B2B sales rep into the field. If the sales rep takes a sales engineer with her, this will entail an additional $2000. Then there’s the reps compensation to factor in. The bottom line is that fully ramped sales reps can cost a company $200,000 – $250,000 a year and more. Therefore, a direct sales force selling $2,000 software is simply not a tenable situation.
Companies selling mid-ranged products usually have inside sales reps that prospect and take orders over the phone. When enterprise-wide deals come up, a specialized sales rep is sent on site. For smaller software packages that vendors want to sell directly, the only solution is to offer the software online and through indirect sales channels such as Amazon and CompUSA.
Charging for Beta Software
Most companies do not charge for the Beta versions of their software. However, some have developed the following approach: Charge the customer for the Beta version and then offer a substantial discount for the final release. The company gets to use the beta version of the product, pays for it and is used as a reference account. When the official version is released, the customer will then “buy” it. How long the reduced price is offered once the product has been released is a matter of negotiation.
What justifies paying for a Beta version of software Consider the value of being a Beta customer. Customers get:
- A competitive advantage gained by accessing a capability before the competitors do.
- The ability to influence the vendor.
- Improved product support.
If a contract calls for onsite training and service to a customer as part of the deal, the revenue cannot be booked until the service is delivered. A simple e-mail from a sales rep stating that they will send someone onsite to help with the implementation can cause problems since it would be considered as an element of the transaction and the revenue would have to be deferred until the company delivered the service.
Make sure the product road map doesn’t end up in the contract. If the contract mentions future product features or that it will support a new operating system, the revenue cannot be booked until those capabilities are available in the product. 
Prospects abhor software whose price is not capped. If a product or a service have variable elements to them, customers will be concerned about what the final cost will be. No one wants uncertainties in their budget so prospects will appreciate a cap to the variable costs. Imagine the first-time cell phone customer who receives a $200 usage bill instead of the expected $70. One exception to the rule seems to be online conferencing applications. Customers are more comfortable with their per usage model because of the conceptual similarity to their telephone service.
Pricing Post Mortem
After setting a pricing model, its always useful to go back and look at the prices that the deals are actually being closed with. In many cases, you will see two things: the average price that deals close at is lower than the list price and a distribution curve of prices. Low closing prices indicate a lack of discipline in the sales force and a price that is set unrealistically high.
When the deal prices fit a wide curve is a negative indicator and can be caused by:
- Lack of sales process discipline. Sales reps are closing deals as they see fit, squeezing as much as they can from customers.
- Customers varying in the value they perceive and therefore the price they are willing to pay.
- Fragmented buying power. The product is being sold into market segments that significantly differ in their ability to pay for a solution.
The pricing model has to be simple. It has to be straightforward for the sales force to calculate and easy for prospects to understand. Before finalizing a pricing model, make sure that a reporter can capture it in a single sentence.
“The fact that our customers probably didn’t understand our licensing as well they might have earlier makes the transition and the perceived pain higher than it actually is — Microsoft’s Steve Ballmer admitting that the shift in Microsoft’s complex volume licensing practices in 1997, had sown some confusion.
Prospects do not like complex pricing models and need to understand the model thoroughly. They need to be able play out what If scenarios when the sales rep is not around. They also need to be able to explain it to others when they are selling the solution inside their organization.
If your company’s name is not IBM, Microsoft or another industry giant, be VERY careful before you create a new pricing model. And if you do try to invent one, make sure you have a fall back plan.
Hubris is one explanation for the sharing savings pricing model. In the 90s NetManage developed a support tool and priced it as a percentage of the software revenues of the customer. Assuming that a company spent 10% of revenue on product support and that NetManages product would save 30% of that, NetManage priced the product at 1% of the customer’s revenue. No one was willing to take the pricing model seriously regardless of the final price point. Customers did not like the model and were not keen on having to have to open their books to a vendor. They felt this pricing model was like a tax. Sales reps were laughed out of multiple accounts. Even though customers are along for the ride, they don’t want to feel that way. Revenue sharing is a legitimate proposal; sharing savings is not.
Understanding pricing is critical to the Product Manager. Pricing goes beyond setting a numerical value for the product. It sets customers expectations, positions the product and impacts the way its sold. Successful pricing is an ongoing effort and should be reexamined continuously as the product goes though its life cycle to ensure congruency between all elements.
This article and its contents copyright (c) 2005 by Daniel Shefer.
Here are a few common pricing models:
Also known as the per seat or named user model in software. This is the way most people buy their material objects: home, car, software licenses, etc.
Cost is determined by the number of users that can access the service, application, etc. at the same time. The concurrent user model is common with server based applications such as databases.
In the per usage model, the cost is proportional to the extent of usage. The most common example is long distance calls and home utilities such as electricity and gas. Depending on the product, an initializing or installation fee might be tied in.
Per Unit of Infrastructure
The product, such as a database, is licensed per the number of CPUs on the machine that runs the application.
The customer pays a percentage of the additional revenue achieved when utilizing the product. It is very difficult to get this model to work. It works best when the vendor manages the collection of the revenue. Some experts warn that they never really work as they are too hard to specify and require too much time to negotiate.
The customer pays a percentage of the savings achieved when utilizing the product. This can cause customer antagonism because the need to open books and share financial information will be seen as an intrusion.
The customer pays a flat fee. Site Licenses are used mostly when usage is wide-spread in large companies. A site license supposedly saves customers the trouble of managing licenses when the number of users fluctuates but is mostly a purchasing ploy. Stay away from it.
Also called the bait and hook or razor and blades model. One product is sold at a low price or even at a loss and a complementary product is sold at a profit to cover the loss associated with the former. Ink jet printers are a classic example. Vendors make their profits on the ink cartridges, not on the printers.
Testing the Validity of the Pricing Model
The pricing model should always be tested against sales scenarios. The best fit should be within the target market. Most models will not be optimized for some segments. In some cases, it may cause money to be left on the table or deals to be lost due to too high of a price. One way to test the fit is to list various sales scenarios and compare the effect on revenue caused by changes of the pricing model and the price points that feed into it. This exercise should be repeated at least twice a year. The assumptions used in the comparison should be validated and the model should be tested on the previous quarters sales.
Another test for the fit of the pricing model and price point within a market segment is that a comparison with the competitors pricing must be made. Take into account the pricing differential based upon positioning and functional differences. If the differences between your price and that of your competitors cannot be justified, you will either have to change the model or the pricing factors in it.
The last test is the market. Make sure that your prospects and customers get it. The pricing model should be simple to explain. If you need more than a couple of sentences to explain the pricing model, it is too complex.
Assuming a pricing model that dictates different prices for each of two volumes of purchases, there are three ways to calculate the actual cost of a purchase Regular and Tiered Volume Discounts and the simpler, all units at nearest tier. The latter is a trivial case so this appendix will focus on the first two.
Regular pricing assumes that each units price between the two tiers lies on a continuum between the two prices. The Tiered approachs starting point is that the total price of the units is a linear line between the total price at each tier point. While these two approaches may sound similar, they produce slightly different results. In fact, Regular Volume Discounts will produce slightly higher revenue than the Tiered approach at any point between the tier boundaries. The equations defining the two approaches are:
X Is the #of units for lower interval.
Y Is the #of units for upper interval.
Z Is the quantity purchased (must be between X and Y).
PX Is the unit price of the lower interval.
PY Is the unit price of the upper interval
To clarify the theory, lets look at a simple example where the lower tier starts at 10 units at $750 a unit and the higher tier starts at 20 units at $500 a unit. The table describes the behavior of the total price for each of the discount schedules:
|# of Units||Cost Per Unit, Regular Volume Discounts||Regular Volume Discounts||Tiered Volume Discounts||Cost Per Unit, Tiered Volume Discounts|
As you can see, while the end points (in this case, 10 and 20 units) are straight forward, the two approaches produce slightly different results in the interim.
Appendix C Variable Pricing
Some of the best-known examples of variable pricing models are eBay, Priceline, the airlines and the stock market. In these markets, the price is set dynamically with little restriction.
eBay is the case exemplar of the ideal marketplace. Geography is not an issue and prices are set by supply and demand. Consumers know what they are looking for and there is no need for anything beyond a spec sheet. eBay might be one of the cheapest selling channels but this type of marketplace is only good for products that are sold As Is with easily defined features. Products that require complex service contracts, professional services or pre-sales are not good candidates for selling on eBay.
Priceline is similar in the bidding aspect but is fundamentally different in another. Technically speaking it is a reverse auction. It adds an uncertainty in the purchase such as the date for the airline tickets or the exact hotel as a trade off for lower prices. This uncertainty can be seen as a limitation of the offered product. In exchange for this limitation and the ability to sell last minute vacancies, vendors are willing to reduce prices.
The nature of these markets limits the type of products that can be sold on them. These pricing model do not fit all products such as perishables. Imagine for example a bakery that offers slightly stale bread from yesterday at a discount. Most people would not be interested in such an offer.
Airlines have made variable pricing into an art. By segmenting their market, they have created a complex, confusing model. They sell a basically identical product at different prices all depending on when you buy and who they think you are. The first parameter is the time of purchase. The price of ticket is significantly lower if you purchase it at least two weeks ahead of time. Airlines assume that anyone buying a ticket at the last moment or that is not staying over the weekend is a businessman and therefore, they can change more. The reasoning here is that the cost will be covered by the employer so the customer is not as price sensitive as those who are paying out of their own pocket. The second parameter is the Saturday night stay over. Airlines assume that business people want to return home for the weekend and will charge more for itineraries that do not include the next Sunday. This confusing approach also infuriates the airlines’ customers, particularly the business traveler.
Appendix D Tips on Collecting Competitor Pricing Information
- Buy your competitors product.
- Have a third party conduct a price survey and call your competitors customers.
- Get friendly with an existing user that is interested.
- Remember that for enterprise products, list prices are rarely the price paid at the end of the day.
- Regularly interrogate the sales reps. Sales reps tend to report competitors pricing on sales that they lost, not the ones they won.
- Add a field for competitors pricing into the sales automation tool report.
- If you meet a competitor at a show, they will probably not volunteer their own pricing but feel free to ask them about a mutual competitor.
- Support Pricing & Negotiating Strategies. By: Fran篩se Tourniaire, The Association of Support Professionals
- Trends in Fee Based Support. By: The Association of Support Professionals.
- Discounting Practices in the Software Industry – A Study of Discounting Use, Monitoring, and Control. By: Jim Geisman, MarketShare.
- Pricing. By: Blois, Keith, Gijsbrechts, Els, Campo, Katia, Oxford Textbook of Marketing, 2000.
- Competing With Gray Markets. By: Antia, Kersi D., Bergen, Mark, Dutta, Shantanu, MIT Sloan Management Review, Fall 2004, Vol. 46, Issue 1.
- When Gray Markets Can Help. MIT Sloan Management Review, Fall2004, Vol. 46 Issue 1, p66, 2p.
- New Rules on the Road toRevenue Recognition SoftLetter Oct, 2004 Vol. 20, No. 23.
- Power Pricing. By: Robert J. Dolan & Hermann Simon.
- The Strategy and Tactics of Pricing. By: Thomas Nagle & R. Holden.
- Strategic Bundling of Products and Prices: A New Synthesis for Marketing. By: Tellis, Gerard J., Journal of Marketing, 00222429, Jan2002, Vol. 66, Issue 1
- Bundling-New Products, New Markets, Low risk. By:Eppen, Gary D., Sloan Management Review, Summer 91, Vol. 32 Issue 4
- Beyond the Many Faces of Price: An Integration of Pricing Strategies. By: Tellis, Gerard J., Journal of Marketing Vol. 50 (October 1986), 146 160.
- Does Promotional Pricing Grow Future BusinessBy: Anderson and Simester,MIT Sloan Management Review, Reprint 45401; Summer 2004, Vol. 45.
- Ipswitch’s Mail customers seem to be in open revolt over a new pricing scheme. By: Joel Spolsky.
- The Role of Rebates in Market Transformation: Friend or Foe By: Michael Gibbs & Jeanne C. Townend, JCF Consulting.
- Rebate Roulette. By: Rae-Dupree, Janet, Spring, Tom, PC World, Mar2004, Vol. 22, Issue 3.
- How to Fight a Price War. By: Akshay R. Rao; Mark E. Bergen; Scott Davis, Harvard Business Review Article, March April, 2000.
- Pricing New Products. By: MICHAEL V. MARN, ERIC V. ROEGNER, AND CRAIG C. ZAWADA, The McKinsey Quarterly, 2003 Number 3
- Articles on the SoftwarePricing website.
- SoftwareCEO.com forum on pricing and licensing.
About Daniel Shefer
Daniel Shefer is Director of Product Architecture at SanDisk. Daniel is a Product Management and Product Marketing professional with nine years of Product experience in the software industry. Prior to joining SanDisk, he was instrumental in product design with Interwise, establishing and managing the product management and the product marketing operations. Additionally, he has also managed technical sales into Microsoft, Avande and others along with driving technology partnerships with 3rd party vendors such as PeopleSoft and Docent.