How To Monetize Data
Businesses are creating, storing, and organizing more data than ever.
The question is, can they make a significant profit from that data?
In many cases, the answer is yes. Firms have to go through a sequence of steps to do it:
- Determine if the data has value
- Package it
- Market it
- Charge for it
Data Monetization Methods
Here's a growing list of the most common data monetization methods:
Selling data is the most obvious and straightforward way to make money from it.
The first decision point for data sales is to consider whether the sales should be exclusive or not.
- Exclusive - The vendor will only sell a collection of data to one buyer
- Non-exclusive - The vendor will sell a collection of data to multiple buyers
As with most products, exclusivity often commands a premium on the sales price. The question remains, however, whether the number of potential buyers in a non-exclusive sales environment would be high enough to create sufficient profit.
Even if a vendor only engages in exclusive sales contracts, vendors may still be able to sell a collection of data to multiple buyers by splitting data sets into multiple pieces. Each piece can then be sold independently.
For instance, if a vendor had a database of soil information for a given town, and a farmer wanted to know about the soil quality of his own land, the vendor would likely find it far more profitable to sell exclusive access to the data relating to that particular parcel to the farmer.
If the farmer wanted access to his soil information, the data relevant to his specific property would be the ideal form of packaging. It would result in the vast majority of value that the entire data set could provide to him while still allowing the vendor to sell exclusive rights to other buyers for their own properties.
The problem with selling data is that it results in a one-time payment. After selling its data, the vendor has to go back to the drawing board to figure out how to make money all over again, and start hunting for new potential customers.
What if there were a way to collect a stream of income? Wouldn't that be better?
Instead of selling data, firms can lease it either for a limited time, or a limited number of uses.
The leasing approach can help improve the cash-flow of the customer, by reducing their upfront costs. Nevertheless, many customers may balk at the idea of having to pay now and again in the future.
Customers will tend to be more amenable to leasing data when any of the following apply:
- Customers have low assets
- Customers have poor cash flow
- Customers need data to be updated frequently
- Vendors have a monopoly on valuable data
Many companies assume that the people who generate data need to be different from the people who pay companies for that data.
Consider the case of FICO credit scores. The scores reflect the financial habits of many people. The buyers of this data are typically corporations who want to understand how the people spend their money.
Sometimes, however, companies accept data from users, analyze it, and store it for them. Some workout monitors, for instance, will track a user's details, create recommendations, and store the results for a recurring fee.
It's often a great idea to sell users' data back to them, because there are no conflicts of interest. Data vendors who can ensure that users own their own data will be able to establish a level of trust that keeps buyers coming back. Additionally the data may be inexpensive to acquire because all of it flows voluntarily from users at no cost.
Provide a Unified Interface for Data
Sometimes there's data that is readily available from other sources. Sometimes it's even free.
Vendors can still make money from this data.
Sometimes data comes from different sources and, as a result, contains records that are formatted and organized in different ways. As an example, some geographic locations may be stored as street addresses in some data sets and as longitude and latitude coordinates in others.
Being able to take all of this messy data and make it conform to a single standard can prove extremely valuable. In many cases, the work required for a single user to clean the data would be cost prohibitive. Fortunately vendors able to resell cleaned versions to multiple parties can turn such data into viable product lines.
Partner with Other Data Vendors
Data is often valuable on its own, but it's often worth more as it becomes combined with additional data. This is often referred to as data enrichment.
For instance, data containing a list job descriptions may be a lot more useful if it could be combined with a list of job salary ranges.
Once data is enriched, it can then be monetized using one of the other methods described above.
Data Monetization Concerns
Data monetization can be incredibly profitable, but it isn't without risk.
Data vendors should think about the following before considering any sale of data products.
There may be legal concerns that limit the ability of vendors to disseminate data. For instance, HIPAA regulations prevent hospitals from selling non-anonymized patient data on the open market. Depending upon the particulars, however, it may be legal to sell the data after it has been anonymized and sufficiently aggregated. Providing the number of patients treated for a given disease would likely be viewed differently by regulators than providing the specific medical history of a non-anonymized patient.
User Credibility Risks
Data vendors risk alienating those who provide them with data if they aren't upfront and clear with their intentions. It can be a trade-off, whereby the more limitations you place upon yourself, the less valuable the data becomes but the more willing others are to provide data.
Buyer Credibility Risks
There are few data sets that are perfect. There are often records that are incorrect, or are converted to the wrong units, or do not meet intended standards.
The higher the error rate, the more likely that customers will become upset with the vendor.
IP Theft Risk
As data is packaged up and made accessible, there's always a risk that the data will be stolen by third parties or used by purchasers in a way contrary to the terms of the license.
It can prove impossible to prevent such issues entirely, but steps can be taken to limit the likelihood and severity of such risks.
Data can be made available through proxy services, such that direct access to significant quantities of the data is inaccessible. An API, for instance, may allow users to search through millions of records but only return the few dozen that most closely match the user's needs.
Data can also be intentionally adulterated to a limited degree to help spot misuse or unlicensed use of data. A collection of email addresses may, for instance, contain a small number of accounts controlled by the vendor. In this way, the vendor can monitor use of the mailing list by inspecting the messages sent to their accounts.
It's often unclear how much a given data set is worth to potential buyers. In many cases, the value of a data set will be worth far more to one potential customer than another.
For this reason, it is often useful to understand the main drivers of value that customers can extract from a data set. Vendors can then analyze buyers according to their characteristics in order to estimate the ideal price for the data set.
Here are a few examples of characteristics that may be value drivers:
- Customer size
- Level of required access to data
- Type of integration into existing systems
- Frequency of data updates needs
- Technical support levels that are requested
- Accuracy guarantees that are demanded
- Associated categorization metadata that is required
Once costs are incurred or sunk, they become irrelevant to decision-making. At that point, the money has been spent and will not magically come back.
That said, costs are very important to think about before they are incurred. If costs are too high, vendors should reconsider selling data, unless there are significant strategic benefits from the process.
So what are some potential costs that data vendors may have to consider?
- Data collection
- Data scrubbing and cleaning
- Data categorization (metadata creation)
- Data packaging
- Data updating
One interesting point to keep in mind is that two companies can start with the exact same pool of data, but differentiate themselves by investing differently in the factors listed above. Data that is accurate, easy to use, and updated frequently will often command significantly higher prices.
This guide is a good start to the topic, but it is just a start.
If you have data you'd like to monetize, wouldn't it make sense to speak to an expert?
I'm available for data monetization consultations!