The numbers for becoming data-driven from a recent McKinsey Institute survey are compelling: 23 times more likely to acquire new customers, 19 times more likely to be more profitable, and 6 times more likely to retain customers. These kind of results are driving many in senior management in pharma to aim at becoming data driven as an organization.
If you are reading this, you either are somewhere on the journey to being data-driven, or are considering it. The number of big pharma who are not yet data-driven, and are still using the same information data systems as they have for the last 10-20 years are still in the majority. Most of these run various database systems, and can either query the system to produce excel spreadsheets or flat files. Then, the analytics teams attempt to integrate the various different systems data in their internal analyses, as this is still largely the only option currently available to them. Pulling static spreadsheets and flat files from numerous different vendor managed databases over months, is not being data-driven in today’s world. Companies still in this space are in a very dangerous position competitively, as they are going to be left behind by more forward thinking companies. So, if you are still in this position, What are the steps required to achieve a data-driven nivana that addresses your specific business needs?
Step 1: Know what you are aiming to achieve
There are numerous reasons for becoming data-driven. Before attempting to solve your data silo issues to collate all your data into a central repository, the first step on the journey to becoming data-driven, is understanding what you are trying to achieve in doing so. We see many companies rushing to get all their data in one big central lake, or buying various big data implementations or tools, before really analysing what their business needs are, and what solutions they should be looking for. There are many great tools out there, but before investing, understand what you need, and then find the right tools for that purpose, rather than choosing a tool and then retro-fitting your needs to the tool.
Starting with a discrete and measurable objective is always a good first step. So many companies start with a fishing expedition to find interesting insights, and while a certain amount of exploration is good, wallowing around in data without an objective it is not a good starting point.
STEP 2: Access the right data
Bigger may not be better.
Recently a company asked us to combine all their data in all their brands in all their markets and have it ready to do AI analytics on it. If we were a tech company first, we would be very enthusiastic, but we do not recommend rushing into this step, unless you have a very clear business case for doing so.
To collate data from all sources is relatively straightforward for big data engineers, and something we do constantly, but it is not quick and easy.
Therefore, it is far more sensible to consider your strategic business goals, and divide these into priorities and then determine what data is most relevant for each of these goals. Take the data ingestion process as a step-wise one where a discrete business objective is identified, then the relevant data for that objective is identified, cleaned and ingested.
STEP 3: Cleaning your data
Everyone knows the saying ‘Garbage in, Garbage out’.
Nowhere is this more of a problem than in analytics. Dirty data is a massive problem. This contaminated data is an issue for most companies. Many companies are looking for an easy way to integrate their big data. An attractive option appears to be using an existing vendor platform in which one can simply upload data without having to go through a rigorous data cleaning process. This is a mistake.
Data has to be cleaned. Just uploading data into a system will mean it is full of errors (multiple instances of the same thing etc.) and if the data is full of errors, so will the results be. This means commercial and strategic goals will not be achieved. So, it has to be cleaned properly before ingesting, which is the job of a data scientist.
Data quality is an industry-wide, global challenge in Pharma. I remember speaking with someone at Bayer in the US. They recognized this problem was so severe, they divided their analytics team into a data team, whose job it was to source and clean data, and a reporting team.
STEP 4: Getting the right data, in a standardized format, in the same location
The ingestion sources for each data source must be created if you want different types of data correctly ingested and standardized, and this is the job of a big data engineer.
Data ingestion set-up time varies by source but can be as quick as 2 days or as long as several weeks of big engineering time per source. That data ingestion also has to be secured and there are numerous processes for doing this but essentially they involve various double encryption of all data being transferred in or out of the system. Then the data security should be audited to the highest standards for security. Setting up these encryption processes also take time for each ingestion type. So, there is a process that takes both time and costly effort from skilled professionals.
STEP 5: Create the right algorithms
The reason for doing everything you are, is to extract more valuable insights to lead to value for your company and your customer to empower your decision making – right? And the reason for this is to give you a strategic advantage which should be reflected in your real world financial results.
But if you are applying a one-size-fits-all approach to your data analytics, you are not going to achieve the outcomes you are aiming for.
STEP 6: Creating interfaces for the analyses that fit in with the goals and end users daily work.
Without ensuring in your initial strategy that the output and content of these analytics is embedded into how the end user works, it would be pointless doing them. If they are not in sync with the day-to-day business and decisions being made, they can make the end user feel like this is yet another thing they need to add to their ‘to do’ list. The point of these is to empower the end users to take the pain out of their decisions and allow actions that they know with certainty will achieve their goals.
STEP 7: Embed within your organizational culture
Finding ongoing success with being data-driven means embedding it into the culture of the organization to gain a competitive advantage and bring more value to the customers. Once some teams start using it and getting better results than other teams, there should be a way to share these internally to help other teams see that this is a strong way to solve their challenges also. When the teams start to realize how complete a view of their customers and business this type of thing provides, they will start to champion the data-driven culture being created.
Conclusion
The process to become data driven does not happen over-night but clearly making the change and becoming data driven as a company will infuse evidence based decision making and create a much higher performing organization. There are a lot of steps involved but the hard work pays off, especially in dynamic changing environments, as the teams can know what to do confidently to stay competitive. And you do not have to do it all at once. Start small with a discrete project for one brand team, then build from the results gained.
Eularis do these projects constantly and have come across most challenges pharma teams face and can assist them overcome these to make the process smooth and successfully achieve your business goals.
This is taken from an excerpt from the more comprehensive white paper available in our resource center.
Found this article interesting?
To learn more about how Eularis can help you find the best solutions to the challenges faced by healthcare teams, please drop us a note or email the author at abates@eularis.com.