Price optimization is, always has been, and always will be one of the hardest tightrope walks in business.

Pricing is an area where your goal as a business (to increase profits) is directly at odds with some of your customers’ goals (to solve their problems at the lowest cost possible). You’re forever balancing the value each customer generates with the attractiveness of your product, and one wrong move either way could mean losing a lot of money.

That’s why we’re here - to continue our masterclass in the basics of business finances, with a special focus on topics relevant to SaaS companies.

Hence why this post will teach you:

• What price optimization is
• Benefits of price optimization
• Price optimization pitfalls
• B2B vs B2C, products vs services
• How to calculate your cost basis
• How to find the demand-side limit for your prices
• SaaS Extra: Finding your true cloud cost basis

Let’s get started.

What price optimization is

Price optimization is the practice of analyzing customer and market data in order to find the best pricing points for your product or service possible. This means the price which hits the sweet spot between being approachable to as many customers as possible, but also maximizing the profit that you’re earning through each individual sale. It also means the price is high enough to cover your cost to produce your product or service plus enough of your everyday operations.

Think of it like this.

You’re a toymaker who’s well-known for quality products that kids can’t get enough of. We’re keeping it simple and saying that everyone wants to buy a toy from you, with the only remaining barrier between you and those purchases being the price of the toys themselves.

Putting the price too high will cause you to lose many potential customers who aren’t willing to pay what you’re asking. Lowering the price will thus increase your overall profits, because earning, say, $10,000 less from your existing customers through price cuts can be offset by attracting $30,000 more in new sales to people who are willing to pay the slightly lower price.

However, you’re still running a business that is looking to maximize its profits where possible. Lowering prices too much could mean that you fail to have enough money to pay for the wood supplies and factory labor required to build your toys. And even if you price your toys for exactly the cost to produce them, you won’t have any revenue left over to pay for your business operations or new toy development.

While you might think that pricing optimization is more of an art than a science, with modern data collecting techniques and (usually paid access) databases you can take advantage of the wealth of market data available. This means that testing your prices against customer demand doesn’t have to be based on guesswork - you can analyze the current market state and eliminate the ambiguity, thus making it much faster and more accurate to optimize your prices. Additionally, calculating your costs to produce your products and services is a relatively straightforward process, though it does get complicated for SaaS companies.

Throughout the rest of this article, we’ll divide price optimization strategies into cost-side and demand-side perspectives. Typically, you’ll need to analyze your pricing from both directions before arriving at the right solution. 

The topic as a whole is far more complicated than the example we’ve just given. Still, it’s a good way to wrap your head around the initial concept before diving into some of the more nuanced aspects.

For example, the increasing prevalence of machine learning in modern businesses (especially in data analysis) is also being seen in demand-side price optimization strategies, with many companies using machine learning techniques to iterate and improve their pricing models according to data points including:

• Customer demographics (age, gender, etc)
• Time of year
• Location
• Current weather
• Competition
• Historical sales

Do you need to analyze all of this in order to perform price optimization? No. Can it help? Well, that depends on the scale of your operation and the amount of data available to you.

A large-scale company with worldwide sales and tens of thousands of customers which has been operating as such for years should have a wealth of its own data to draw from when analyzing pricing points from both a cost and demand perspective. From its own information alone it should know the exact costs to generate each product and service, and be able to pull together the trends of which elements seem to affect the price their customers are willing to pay.

A small-scale bakery that barely keeps its single location afloat will have much more trouble. Not only will customer data be limited, but as a bakery owner you won’t be able to split out your monthly costs by each of the items you sell. It might sound obvious, but price optimization is less effective the less data you have to base your analysis on.

But enough of the doom and gloom - let’s dive into why you want to be optimizing your price.

Benefits of price optimization

The primary benefit of price optimization that almost everything comes back to is building a reliable margin - or the difference between the revenues made from selling your products and the costs incurred to build them. The ways in which this is done are the elements that you’ll get more or less mileage out of depending on things like your business size, type, and so on.

Most directly, price optimization increases margins by charging customers the greatest increase over the base cost to make your product or service, all while walking the fine line of not turning them away with too high prices. There are other benefits we’ll talk about shortly, but this really is the vital core of the practice.

On one hand, you’re getting prices as high as they can be in order to maximize the profits you’re earning from a single sale. On the other, you’re keeping prices in a realistic range that your target audience will both be able to afford and be fully willing to commit to. It’s a balancing act with the wealth of your business on the line, so it pays to be using data-backed methods to be sure you’re going in the right direction.

If you’re using machine learning to analyze pricing, you also increase your profits by reducing the costs involved with manually analyzing the same data. Sure, you could hire someone to analyze everything manually and inform your future pricing decisions, but this is only realistic in a small-scale business that doesn’t have much data to begin with (and thus stands less to gain from the process in general).

Not to mention that having a consistent approach to price optimization will also make your pricing plan more consistent. This will reduce any negative opinions generated by changing your prices too often or too wildly, but more on those in the next section of this post.

Finally, focusing on price optimization helps you to at least partially automate the entire process, saving you time and money that can be better spent in other areas to increase your profits and productivity further. Once you have an effective analysis model to work from, the most that you will generally have to do in subsequent analyses is plug in the new data that you’ve gathered since the last time you carried it out.

Price optimization pitfalls

Sadly, as with many things, not everything is positive when it comes to price optimization. The issues here mainly stem from the complexity of the topic and the risk of getting things wrong, thus alienating much of your customer base.

First off, price optimization is far more complex than “tweak your prices and see what happens”. As we’ve covered above, there are a huge amount of contributing factors to the topic and a sea of potential data that you can easily drown in if you aren’t experienced with large-scale analysis.

This means that you’ll either need to hire someone who is experienced with data analysis, or leverage a third-party service that can do the heavy lifting for you. Obviously, the latter option will generally be cheaper and faster for a one-time analysis, but hiring an analyst shouldn’t be ruled out if you can make use of them in other data-heavy areas. Plus, once they’re settled in and know how your company works and what its history looks like, a hired analyst will be able to carry out future price optimization much faster than their first time.

Second, carrying out price optimization at all means that you will need to change your prices. This means not only figuring out what they should be changed to, but also how they should be altered and how to present the change to your existing customers.

Remember that your customers aren’t emotionless figures who purchase based on logic alone; if they feel that you’ve insulted them by changing the price that they paid or are currently paying, there’s every chance that that negativity could show itself in bad product or service reviews, or even a general reluctance to continue being a customer. That may be fine if you intend to focus on a completely different buyer or channel for selling your products, but this should always be a consideration and not be done by mistake.

This leads us neatly to the final pitfall, which is that you could lose customers if you don’t deploy your changes carefully and effectively.

It doesn’t matter how much data you analyze or factors you try to take into account if you aren’t also considering the impact that your pricing change will have. What on paper may be a logical increase to bring your prices in line with your product’s worth could be seen as blindsiding your customer base or short-changing those who enabled your company to grow to the point that it’s currently at.

Thankfully, there are ways to mitigate this last issue in particular. All you need to do is to consider ways in which you can provide a benefit to existing customers too (especially if they’re making ongoing payments).

For example, you could offer to allow existing customers to remain on the payment plans that they signed up with, or by offering them a discount when moving to the new plan. If you choose to do the former, it may also be worth making sure that the product they’re using remains in the state that they bought it too - you can encourage them to move to the new pricing plan in order to access new features.

B2B vs B2C, products vs services

Before diving into how to optimize your prices we need to lay the groundwork of how different businesses interact with price optimization. 

B2B, or companies that sell to other companies, and B2C, or those that sell to consumers, businesses differ in the data that’s available to them and (generally) the scale of their pricing plans.

B2C tends to focus on smaller products or services that require relatively minor changes in order to optimize their pricing, and can sometimes be more heavily influenced by aspects such as the time of year for seasonal discounts. Consumers don’t expect fixed pricing, nor does a small price fluctuation affect them too drastically. Meanwhile, B2B will generally offer more expensive products or services to fewer buyers, leading to a much smaller market data set and buyers who expect your costs to fit their budgets precisely to boot.

What your company is selling or providing will also affect how you’ll optimize your pricing. Those selling products will have to reflect more on the cost of aspects like physical materials, manufacturing, and logistics (most of which have been skyrocketing in the last few years) and whether their prices need to be raised to account for these things alone.

Ongoing services will focus more heavily on their pricing in relation to their customers’ lifetime value (LTV), as even a minor pricing change can quickly compound over the many payments which make up a single customer’s lifetime with the company.

Further, the age of your company and its funding can result in a focus more towards either the cost or the demand side of pricing optimization. For example, a young, venture-backed company, like Uber in the 2010s, doesn’t have any concerns about having the cash to pay for company operations and can price their products and services at cost. This will eliminate any margin whatsoever but in return will promote rapid growth in market share. On the contrary, a centuries-old public company, like General Motors, needs to bring in enough revenue on their products to cover not only the cost to manufacture but also all new product development and administrative overhead.

All of these elements and more need to be considered when thinking about how to optimize your prices. Let's start with an outline of both the cost-side and demand-side frameworks for price optimization.

How to calculate your cost basis

Before we can consider how much your customers might be willing to pay for your product or service, it’s imperative to know how much it costs to deliver it. Essentially, this is the study of unit economics and while it can be difficult to accomplish, the process is rather straightforward.

First, consider what each unique unit of purchase is that will be considered for pricing optimization. It might be a loaf of bread, as in the bakery example, or it could be an Uber ride. For a single-product SaaS company, this could be a customer while for a multi-product SaaS company this could be a product subscription.

Next, identify all the costs required to generate your products and services for a given time period. For manufactured products, like a General Motors car, this includes everything from raw material costs to assembly labor and factory electricity. For professional services, this may simply entail the salary of your employees and their travel to each work site.

Finally, isolate these costs down to the individual unit of purchase. In the bakery example, this means dividing up the cost of flour between the bread loaves, the muffins and the cakes, proportionally to the amount used. Then you’ll ass that to the cost of labor divided proportionally by the time it takes to create each product.

When you have finished this exercise, every expense that your company incurs to deliver a product or service should either be categorized directly as associated with a single unit of purchase, or will be allocated by proportions according to a logical rule. Your financial team will be able to report the exact cost to generate each unit, precisely.

Once you know your costs, you can start to analyze your price points.

How to find the demand-side limit for your prices

In contrast to the cost side, there is a lot of nuance in how you go about price optimization on the demand side. We like to break down the process into four simple steps:

• Gather customer data
• Define your value
• Analyze your data
• Test pricing changes

One of the most complicated actions comes first, in which you need to gather as much customer data as you can. If there are market databases that you can access, do so. If you have historical data from your own company, even better. Customer reviews and the time they occurred, support ticket feedback, net promoter scores, previous pricing plans and discounts run, MRR, LTV, churn rate, market trends and size, you name it; if it could be relevant to your pricing in any way, you need to be gathering that data to analyze it later.

While you’re busy gathering that data, you should also be looking long and hard at your products and/or services and defining what exactly their value is to your customers. From this you can determine whether your pricing needs to be structurally changed rather than just editing the values.

For example, as a SaaS company you may realize that your previous pricing was based on how much it cost to create process documents on your servers, when your value could actually be based on how many times your clients avoided an audit due to following the process documents, because that is what is most valuable about your software to your customers. With the cost of an audit reaching into the high five- or six-figures, that’s quite a significant customer value.

Analyzing your data in the context of your value will then allow you to see trends in how previous pricing changes affected your sales rates, and how you can try changing your current prices to better optimize them. This is where you really need to utilize a data analyst or third-party service to get the best insight into your data. Analysts often find that, contrary to sales organizations’ belief, lower prices don’t always drive customers to purchase. However, once prices reach a certain threshold, they do add friction to getting the deal to close.

Finally, as with almost everything else, you should never deploy your pricing changes without first testing them out. Your approach to testing will depend on your volume of sales. In a high volume scenario, testing should ideally take the form of A/B testing your pricing with a range of customers (particularly your core audience). In a low volume scenario, an ideal test should involve qualitative customer research interviews that ask customers to reflect on their willingness to pay for certain product or service characteristics, either in isolation or in a bundle.

Remember, as with all research and testing, don’t let your focus on pricing and its effects rule out the possibility of other factors driving purchase decisions. You don’t want all of your hard work to go down the drain at the last minute due to an unaccounted element skewing your results without you realizing it.

SaaS Extra: Finding your true cloud cost basis

In the world of SaaS, the primary cost to deliver your software is your cloud bill. No matter if you’re in a colocated data center or fully cloud-native on AWS, it’s the cost of cloud that determines the lower limit of your pricing.

But there’s just one problem; the cost to deliver a single product experience to a single customer bears no relationship to the rent of a single server. So how do you go about identifying your true cloud cost basis?

There are a variety of factors at play that will determine your best approach. These include tenancy structure, product isolation, and customer profile segmentation.

For a SaaS business serving enterprise customers who are each in an isolated single-tenant environment, the cost to deliver your software to each customer is unique and easily quantifiable. In this case, the cost basis per customer may be drastically different and so too should be the price analysis. On the contrary, a multi-tenant environment requires a deeper look at the other two factors.

In the event that you sell more than one product that can operate independently of the other products you sell (as opposed to add-ons or additional features), it is important to treat each as a unique unit of purchase. In some cases you may have isolated these products into separate AWS accounts, while in others the only way to tease them apart are through namespaces in shared Kubernetes clusters. No matter the methodology, you will need to identify and separate each.

When your customers all fall into a specific customer profile, such as K-12 public school districts with no more than 100 school sites, it can be easy to take your total cloud cost and divide it evenly across your customer count to find your cost basis. If this is not the case, you may need to create isolated customer profile segments, estimate these segments’ relative cloud cost in comparison to the other segments, and then weight the total customer counts in each segment by these comparative values.

But, most importantly, this information must be available to the finance team in the accounting system. Without having an accurate recording of the cost basis of every unit of purchase, it’s impossible to track your margins and chart your course to profitability.

Luckily, there’s a solution.

With Aimably’s AWS Invoice Management Software, you can easily categorize and allocate your cloud costs and deliver these calculations directly into your accounting system alongside your bill at the end of each month. It’s the easiest way to make your AWS accounts work with your accounting system, instead of making things far more complicated.

There you have it! From here on out it’s a case of repeating the entire process every so often (depending on your other priorities and how often you introduce new products or customer segments) and iterating to continually test and improve your pricing. Even if you don’t find a price that tests better than your current one, at least you can sit safely in knowing that you’re putting your tactics through their paces to stay at the top of your game.