The revenue model for a digital service needs to cover the costs of running it
Making a smart product needs more than adding an app. You need to cover the costs of the digital service or risk losing money. Which revenue model can you use? And why would anyone pay for this service?
In the B2B market, companies know that services are not free and are used to maintenance contracts. In the B2C market, consumers are used to free services that are paid for by advertisers or other parties. Let’s consider the options for a new B2C product, for example a smart refrigerator.
What are the costs of providing a digital service? To be able to run the service, you need to add hardware and software to the product. On the hardware side, you could add sensors to measure how the fridge is functioning, chips to enable wireless connectivity, and a physical control panel. On the software side, you could add control software for the panel, a web-based interface, and an app to interact with the fridge.
These components bring with them additional costs that you need to cover. The hardware costs and upfront software development costs can be covered by a higher price for the fridge. It has extra features, so a higher price is expected. So far, the revenue model is straightforward.
The digital service costs are more complicated because they continue after the sale. They include providing feature and security updates, storing and analysing customer data, providing customer service, and so on. You can estimate for how long the service needs to keep working, then add X years of its costs to the purchase price.
Using new sales to pay for the service
If the service needs to keep working “indefinitely” then you will reach a point where the incremental costs of maintaining it have exceeded the revenue you earned at the time of sale. Thus, there is a potential mismatch in both the amount and the timing of the revenues versus costs. If you do not balance these, the service will be loss-making for the company.
While sales are growing, you can cover the costs through new sales, subsidising the costs for earlier products that are still using the service. This delays the moment when the service becomes loss-making but does not prevent it. If your service costs $10M a year to maintain, and your uplift in product price is $10, then selling 1M products a year will cover it, as long as you can make those sales each year. This model also needs the incremental cost of serving an additional customer to be relatively low.
This “new sales to cover costs” model is the reason why software makers keep making new versions of their products to sell (e.g. MS Office, Windows). Once everyone has the product already, you need to give them a reason to spend more money buying a new one. The model stops working when the product they have is good enough (e.g. Windows XP, then Windows 7), which is why software makers have switched to the subscription model (e.g. Office 365, Adobe Creative Suite).
Avoid having to shut down the service
If revenues from sales are insufficient, then you need to find a new revenue stream or shut down the service. Computer game publishers regularly shut down the multiplayer servers of computer games a few years after they have been released. Since there are no incremental revenues, the publishers cannot keep up the incremental costs. From then on, the games can be played in single player mode, or not at all, so they are partially or completely non-functional. The result is a time limit on the usefulness of the product.
In the case of the smart fridge, it would be difficult to justify to customers that the digital service is going to stop working before the lifetime of the fridge ends. This is even more true when the service is the core functionality of the product (e.g. a Nest thermostat). So this naturally leads to the question, “What new revenue streams can we generate with the digital service?” in order to cover the ongoing costs.
Make a service that meets customer needs
We started from the product side with the idea of a “smart fridge”. It’s an easy place to start but it misses the other side of the proposition, namely the customer. Instead of asking, “How can we get someone to pay for this smart fridge we invented?”, we should ask, “What customer need is there that the smart fridge can satisfy, and why would anyone pay for it?”
Let’s suppose that the customer need is to plan meals around the supplies in the fridge, and restock items that are finished. The service could be to detect the food in the fridge and provide shopping lists and recipes for family dinners. The customer need determines the fridge design. On the hardware side, the fridge will need cameras to see what is inside. The software could come with a recommendation engine for recipes, a learning algorithm for what the family usually eats, and a database of products to be tracked. Over time it could aggregate data from all its users and stock available in local stores and start to recommend new options for the family to try.
Eight possible revenue models
What revenue models could work for this service? Here are some options.
- Single fixed price: Pay once at purchase. We discussed this option above. It’s the simplest and most common, but will not cover the ongoing costs.
- Pay as you go: Each time someone uses the app to create a recommendation, a small charge is incurred and a bill is sent periodically. This is how digital parking services work, or hailing a cab, or renting a movie. It works well when an instance of “use” is well-defined and easily metered. It can work, but it discourages playing around with the service and requires detailed administration to track and bill usage.
- Subscription: Pay a fixed amount periodically. The most common model for selling software, services and access to digital content. From the supplier point of view this is easy to administer. From a consumer point of view, it is familiar, and people tend to use it for services that they value. Will they value this service enough to pay a subscription? Can you charge $5 a month, or $30 a year? Consumers will compare the service to other things that cost as much, consider how often they use the service, and if they can get the same for free elsewhere. Free substitutes are the downfall of many subscription models.
- Time-limited fixed price: The price upon purchase includes service for X years, and that extending the service for another X years will require a new payment at that time. This is a variant of the “extended warranty” model, except that you are not asking for payment upfront. This model would mitigate the risk of lifetime costs exceeding revenues. The downside is that if users do not extend after the first X years, you end up with a smaller paying user base to cover the costs of keeping the service running. The upside is that if some users do pay, you end up with more revenues than the single fixed price model.
- Freemium: Some features are free and the rest require a payment (either one-off, pay as you go or subscription). This is another common revenue model in software (e.g. Evernote). It is an asymmetric model where a small group of paying users subsidises the service for the free users. In some cases, the free users provide data and a community that is also valuable to the paying users. This model could be applicable if some features are so desirable that consumer will be willing to upgrade.
- Tiered pricing: Different levels of features are provided at different price points, to different customer segments. This is a variant of the freemium model. This model is common in business software, with “home”, “small business”, “enterprise” variants (e.g. Microsoft Office, Salesforce). Given that we are serving individual consumers, this model is not relevant in this case.
- Advertising: Ads are displayed on the screen or in the app. This model is common on the internet, on social media and in phone apps. It requires participation in an ad network and the ability to provide a well-described demographic for advertisers. In this case, it could work for food ads from retailers. Consumers might be surprised to see ads on a premium priced product that they paid for already. They associate ads with free products: the value exchange is that they accept the ads for free service, and they know that someone else is covering the costs. If they already paid for the product, they believe that they have already covered the costs. So consumer acceptance could be an issue in this case.
- Transaction fees: If the consumer buys a product through the service, you earn a commission on the transaction. This is a platform play, where the win-win for you and the retailer is that the consumer buys more food there, so the retailer is willing to share the benefit with you. The consumer may be charged a transaction fee, e.g. the same fee that is charged for home delivery by the same retailer. A promising revenue model, that requires building a platform and setting up collaborations.
You need to beat “free”
No matter what model you choose, your competitor is the free internet that everyone can access on their phone. If you’re providing information, they can Google it. For e-commerce, they can go to an online retailer. If the service works by analysing pictures, someone can build a phone app that uses pictures of a non-smart fridge to provide the exact same service. If you’re building a community of like-minded individuals, users can find a group on Facebook.
Your offering needs to be better than free substitutes by providing unique features and being specific to your customer and product. Physical sensors in the fridge enable you to collect data that cannot be replicated with a smartphone, and service around that data (like fridge maintenance or food storage advice) could be more immune to substitution.
Making the choice
So which revenue model do you choose? If this is your first entry into digital service, don’t make it too complicated. Despite the downsides, single fixed price is very common (e.g. Nest, Philips Hue, Sonos, GoPro) as well as fixed price with premium subscription (e.g. Fitbit FitStar). Time-limited fixed price could work, but don’t know of any examples yet. If you do have an example, please let me know.
For a real-life example, Samsung’s smart fridge sells at a significant premium, so its main revenue model is single fixed price. It is running a platform with a wide range of partners and apps that are free. There is a shopping app with MasterCard that could be earning transaction fees. For Samsung, the incremental cost of a few fridge apps is tiny compared to its existing resources for its phones apps, so it may not be worried about covering the costs of the fridge apps.
If you want to make a digital service, make sure you get the value exchange right, and be clear which revenue model you are going for. Ask yourself two questions: “Why would anyone want to pay for this service?” and “How can my revenues cover my ongoing costs?”
© 2017 Veridia Consulting