Understanding the Dynamics of the Modern Data Buying Process
Buying information – especially financial data – has historically been an arduous and expensive process. Arduous because data is complex and in most cases the people doing the buying are likely not the people who will be consuming the data, leading to a lot of back and forth, misunderstandings and wasted time. And expensive because, well, data doesn’t come cheap, and the complexity of licensing agreements requires expert attention.
But the explosion in new data types – alternative data, unstructured data – and the ability to slice and dice existing ones into more granular micro-data services, is changing the data buying process. Increasingly, through the use of data sharing and data marketplace platforms, data providers are offering more control to the ultimate consumers of their information services. The emerging picture is of a more streamlined data buying and selling interaction, with well-defined roles for different players in the process, which itself is characterized by recognizable phases.
Before looking at the key steps of today’s data buying process, its worth considering the key players. First, there is the data provider or seller, that is the entity seeking to sell the data it has generated from its business activities, or makes its living from aggregating third-party information to create a data product. Second is the data buyer or consumer, an entity or individual that seeks to buy data it requires to satisfy some requirement or other. Finally, there is the end-user, the individual who will actually use the data in question; this person may or may not be the data buyer in any given transaction.
Under the scenario outlined above, the modern data buying process is characterized by three distinguishable phases: Discovery, procurement and access/distribution.
The Data Discovery Process
For any potential buyer, the data discovery process is sparked by some form of question that needs to be answered using data. This may be a financial trading decision, or research for product development, or an enquiry about product inventory. In reality, any kind of business decision usually requires the use of data in some way during the decision-making process. Whatever the reason for the enquiry, the data buyer is almost certainly looking for data that is most relevant to his or her organization’s needs.
In the traditional data buying model, buyers are hugely reliant on sellers for information about the available data sets or products, and this is why this phase is a major point of failure for many data buyers. Data sellers often aren’t familiar with the nuances of their own data, and how it might be applied to specific use-cases. This is why the modern data buying process typically uses a self-service model that empowers buyers to identify the data that best suits their needs.
In our experience, there are two types of persona that data buyers adopt during the data discovery process. The first of these is the Hunter persona. The Hunter knows exactly what they want and is seeking to get it in the best possible form available. Examples may include tick histories for a specific financial instrument traded on a particular exchange over a defined time period. This data set may be available from a variety of sources, and the Hunter will want to understand quickly the different characteristics of each supplier’s data sets, including time span, granularity of time-stamp, data format and so on.
The second type of persona is the Explorer, who as the name suggests wants to know what’s available even if they don’t have a specific idea of what they require. The Explorer is inclined to browse and research in order to identify which data set might be useful for answering that original business question. More often than not, the Explorer is investigating data sets on behalf of an end-user, although that’s not exclusively the case. In large organizations, the Explorer may be part of a sophisticated procurement team. They may use the services of another persona – the independent third-party data broker – although as the data transaction becomes more streamlined and self-service, the need for independent help in data sourcing and buying is likely to recede.
For each type of buyer persona, the buying process can be painful, with the right data difficult to source, hence the existence of the data broker under current models of buying and selling. Again, self-service data discovery features help both Hunters and Explorers get the information they need quickly to make a buying decision.
Once the data buyer has identified what they want to buy, the procurement process is triggered. This is the most nuanced aspect of the data buying process, due to the complexity and diversity of supplier licensing policies. To navigate this process, most data buyers make use of two key roles: those of Controller and Consultant. The Controller is the executive at the buying organization charged with buying the data. This could be a dedicated procurement executive, or it could be someone closer to the end-user, like a data scientist. In either case, this person will interact closely with all relevant players within their organization, including the end-user.
The Consultant is a person or organization engaged in the buying process only temporarily. This could be a legal team brought in to review contracts, the procurement team handling money orders and other logistical issues, or a senior executive within the buying organization. Each of these Consultants has a different role with different interests. The legal team, for example, will look to red-line data licensing contracts, but would not be involved with negotiating terms of the contracts. In the modern data buying process, much of this interaction between Consultants and the data seller is handled online, with the seller often responsible for providing tools to make this possible.
Data Access & Distribution
The final phase of the process is data access and distribution. Here, a key persona is the data engineer, who is responsible for taking the data – whether via FTP, API or streaming – to create pipelines that will deliver the information to the end-user. This may require some form of data repository to ensure the data is delivered in a state and format that is usable by the end-user. Invariably, in the traditional data buying model, the data engineer is not the person who will consume the data but rather the conduit for ensuring delivery to its final destination.
New data marketplace and self-service technologies increasingly allow the end-user to take on responsibility and control of this process from start to finish. In this emerging model, it is the end-user who performs data discovery, drawing on descriptions and trial versions of granular data sets made available through data marketplace platforms like TickSmith’s Enterprise Data Web Store. It is the end-user who assumes the role of Controller, with the authority to pass contracts to appropriate Consultants like legal teams where needed. And it’s the end-user who becomes the data engineer, by accessing the data through an easy-to-use access mechanism supported by the modern data marketplace platform.
This modern approach to the data buying process allows data sellers to offer control to the end-user, through self-service methodologies. In many ways, this is made possible by the ability to create more granular data sets for commercial distribution, which may bring down the cost of the specific data any given end-user is seeking by circumventing the data packaging approach used by traditional data suppliers.