‘Not having a big data strategy can hurt your business big time’ In an exclusive interaction with Express Pharma, Simon Ulrich, Head of B

What is the business case for a pharmaceutical company to invest in a BI and MDM strategy? What are the primary challenges you see in pharma industry which can be solved using BI and analytics?

Hoffmann-La Roche

A challenging question! The primary challenges we face in pharma industries are at one front very similar to other industries such as to optimise our supply chain, generate accurate financial reports, etc. but on the other hand, pharma faces few unique challenges on its own such as, understanding the influence network around a decision if an innovative drug will be made available for treating a patient.

This includes physicians, healthcare providers but also payers (e.g. insurances) and patients themselves. As we sell and distribute drugs mostly to an apothecary or wholesalers and not directly to healthcare providers, physicians or patients, the challenge is to understand whether influencers are aware of our latest drugs and are using it for optimal treatment. Analytics play a very important role here to help us understand the influence networks at least on a macro level and identifying places where we need to raise awareness as well as guiding us to use marketing effectively.

If you observe how the healthcare market is changing, there are two new non-prescribe influencers added in today’s era. On one hand, the payer or insurance carrier plays an important role in deciding which drugs should be prescribed by providers from a cost effectiveness point of view. And on the other hand, individual patients who are tech savvy and participate in review forums, micro blogging etc. are becoming more popular influencers. Knowing these details and a 360 degree view of patients is becoming extremely important to understand and big data and analytics play another important role here. While part of this information comes from internal sources, to acquire this picture we have to work with (often multiple) external data providers. This generates a significant integration challenge and business teams are right to demand state-of-the-art BI applications and services to help them tackle them. While I am happy to see new opportunities arising from health IT, this of course generates significant challenges from an informatics perspective to make sure they can be converted into reality. Analytics play a key role here.

Why is data analytics becoming so crucial in the pharma industry?

I would argue that it has been crucial for the industry since quite a while, for reasons discussed in the previous question. We also have to keep in mind that the market is actually getting significantly tougher. Government increases regulations to comply with. It gets more difficult to develop new blockbuster products keeping in mind that one needs to invest more and more in research to get there. Further genericisation is making it harder to achieve ROI with your products. Analytics of course can’t solve that problem of the industry. I see it rather as a tool that is being applied intelligently which helps to address upcoming challenges more effectively.

How do companies like Roche determine ROI on such spends?

It is another difficult question to answer. Naturally it is hard to proof the exact financial value of the often quoted “better decision making.” Honestly, we do have analytics projects where we can’t calculate ROI. We know we need them either to gain insights (faster) that will generate value or they are mandatory to track success of business processes and changes. In those cases we normally assume that part of the generated value is to be attributed to the analytics implementation/service that was part of enabling it. There are also the more easy cases (which are getting less and less though) where analytics replace manual work processes either in business or IT so ROI can be calculated based on savings generated by that.

What in your opinion should be the road map to a company devising and then implementing such a strategy?

You should be closely engaging with your business. You should ask questions like, what questions do they need to get insights on? How should information be accessible within your company (get some external input as well). Define your vision as where you need to be in next two-three years. Assess your gaps versus this vision from a technology and organisational readiness standpoint. Devise your hedgehog principle, what your teams really have the potential to be best at, for your organisation and be clear on how you would like to structure yourself. Decide which companies you like to partner with to cover the areas where your internal team cannot be best at. While I tried different strategies in the past I have found that it is of higher value for a non IT technology company to attract talent focusing on business understanding, project/process management and doing the triage between business output and it service. Pure play technical skills are easy to buy from the IT services market compared to those skills.

Last but not the least, do the back planning to establish what your teams should focus on today.

Should such a strategy include external BI vendors and if so, given the knowledge driven nature of the industry and the attendant security concerns, how would you determine the reliability of such a partner?

Yes, I think that is mandatory. Especially, as in BI the demand for solutions is highly fluctuating. Without strong external partners you kill your organisation by constant hiring and firing. I strongly believe every manager should focus his team on what potential they have to do best for the company. If that is not enough to provide the needed results then you should partner with those organisations that have this ability/potential to provide. Now assuming the “basics” are there, the company I would like to partner with should have the ability to pass security audits. Their rates should be in a range to be cost effective. I determine reliability to be of prime importance by the performance delivered by every individual my team is working with. I know that with that I am going against the trend of managed services but I believe that all services are in the end provided by people. So having the best people is the building block to have the best service delivered. Secondly, seeing that our partners are taking accountability for issues and helping us to develop new business opportunities is critical to establish a long term trust based relationship. Having said that , I would never rely on a single supplier only. The risk of that would just be too high.

Is MDM a must or optional for mid-sized and/or niche pharma industry going forwards?

I have to say that I worked all my life for big pharma companies so I would like to put a caveat before giving an answer. MDM is becoming mandatory irrespective of company sizes there are many legal regulatory requirements that a company needs to fulfill . In addition, sooner or later a small company needs to partner with a big company for growth and having a clean MDM is a very essential component on establishing reliable partner.

The pharma industry generates a lot of data. What in your opinion can be the potential effective business cases for big data?

I see multiple business cases here. Research is already working with technology that got wrapped under the term “Big Data”. If you just think of the volume aspect out of the 3V (Volume, Velocity and Variety) and think about a genome sequencer which with one trial run can easily generate 6 PB of data. So if you need to analyse that then you need to find the best cost effective ways to store those data and run statistical models to get the best meaningful analytics done on it. Another example would be to capture machine generated data on manufacturing units for batch specific data and predicting batch failure before it occurs to control cost more effectively. Also the classics of sentiment analysis can be an interesting business case as for most of other industries as well.

Another case for the pharma industry can be identifying adverse events from patient sentiments. Getting even more input to be considered for improving drug quality. Big data plays a quite important role to get that to us. As mentioned before, health IT and the challenge to get a 360 degree view of your customers is another important business case of big data. Strong statistical analysis can help you to better manage your trials and proof superior value of your products. Lastly I think expert systems will become more important in the healthcare industry as a whole in the future. Ultimately they (can) rely on several sources of “Big Data.”

What are the few things the CIO of a pharma / life sciences company should look out for while implementing BI and MDM strategies?

Not that I have been a CIO before but from a BI technology prospective, I would recommend CIOs to look for,

Mobile BI solutions – To generate device independence data.

Contextual/ operational BI – detaching BI from data warehousing system but. integrating them with transactional systems to make operational decisions in real time.

In-memory DBs – to meet the ever growing demand for speed and faster decision making

Big data – A pharma company can’t ignore big data in their strategies. It is a must-have business tool and if ignored, you can hurt your business big time.