Artificial Intelligence is on the C Suite Agenda – but Ineffectively Implemented

It is good to finally see that big data and Artificial Intelligence analytics are now on the pharma C-Suite agenda with a recent survey by NewVantage Partners showing that 85.5% of major corporations are working on big data initiatives.  This is because many companies have now measured and proven the value of these for growing their pharma brands and companies.


Although many in pharma C-suites want to utilize big data and AI, in many cases, they do not really understand what this means and how it can be used effectively to grow their company. It is important that they understand how their company can effectively utilize it, as the impact of this is profoundly changing the dynamics of the pharmaceutical industry in areas from R&D, to manufacturing to supply chain to sales and marketing.

Some companies are utilizing this to improve revenue, innovate, increase customer engagement, decrease costs, and transform the business. However, many who are attempting to do this are reporting organizational impediments. Some of these impediments include lack of IT infrastructure, lack of access to their own data, roadblocks from vendors who store various parts of their data, lack of understanding how to proceed and lack of adoption by the end users in the company. All of these are solvable but it is becoming clear that organizations need to adapt in many areas before achieving optimal transformation.

We are already at the stage in which we can use AI to analyze real time, and real world, data automatically and automatically utilize the outputs to achieve transformational growth but are we? How can your company transform their ability to implement and reap real transformational growth from AI?

1. Ensure the Business leaders understand the basics of AI and what is possible
In many cases, CEOs are on board with the need, but pass responsibility to a team to implement. They are busy and this is not a core competence so the approach is understandable. Sometimes this gets passed to a commercial team, sometimes IT, sometimes marketing, and sometimes straight to data scientists, and in some companies, a multi-disciplinary team (which is better). And sometimes one internal team hands it off to another. We have seen marketing teams who have been given responsibility go to the internal IT teams and ask them to do something as they are unsure what to do. This is a waste of valuable people. The leaders need to understand what is possible with big data and AI in the various areas of the business and plan resources appropriately.

2. Refresh the mind set
Teams – from the C-suite down – need to all consider what their challenges are in their jobs to be done. In fact, considering the ‘job to be done’ is often the best way to approach this as much disruption has been created from this view. By continuing to look at the challenge in the same way as many do now, one may use AI to create efficiencies but using AI and big data should not simply be about automating business processes but about automating thought processes to create far stronger businesses. Henry Ford once said ‘If I asked people what they wanted, they would have asked for faster horses.’ By considering the ‘job to be done’ we often find better solutions.

3. Planning a strategy
Everything needs a strategy to succeed. Without a clear strategy and well planned initiatives to achieve the strategy, with measurement of success built in, it will be difficult to achieve the success required. The teams need to establish their strategic priorities, and plan in much the same way all marketing and other priorities are planned. It could be simple things such as speeding up clinical trials by collecting real time data from clinical trials automatically, or create a winning launch for a drug, or simply help the sales team become more effective in their calls to enhance their results or many other areas. The big data and AI is simply the enabler that allows far strong results than ever before as it takes into account more data and stronger modelling than ever before. The planning requires someone with an understanding of data and AI and what is possible and what is not, AND someone responsible for the strategy. The 2 areas must be synthesized in a collaborative way in order to get good results. Without strategy you would get a host of siloed initiatives for specific AI products that may be great but may not actually be solving the strategic problem at hand.

4. Planning data access
Awareness of what data is available internally, along with access to internal data, can become a roadblock to project schedules. We have found our client teams run into so many internal road blocks in this area. There should be someone in the company who can tell teams what data they have, what detail is in the data, and who can access all the relevant data What is interesting is that so many companies have their data stored on different vendor databases. And, many vendor software or database companies, despite only being where the data sits (the data itself is data the company entered, e.g. CRM data) do not allow their clients own data to be used with anyone apart from them. One top 10 pharma company is spending 3-4 years and many hundreds of millions replacing all their legacy systems with cloud infrastructure and all their data in all countries will go to their data lake and be able to be modelled in any way they need. This is a big project but ultimately one that will mean that any data they need for any strategy, in any country, they can access immediately, anywhere. Pharma companies do need to start to take control of their own data to avoid these types of situations where they are held hostage over access to their own data.

5. Securing expertise
In any big data and AI project, you will need access to expertise. The resource demands are large from the strategy planning to the big data IT infrastructure set up, the data scientists and the visualization and programming, not to mention the actions to be taken by the teams also. There are many tools available for many of the components of this process but there is no one-size-fits-all approach. There are numerous companies selling parts or all of this process and the teams need to analyze (or get help analyzing) which should be building themselves and which parts they can use existing tools. A buy versus build analysis should be done for the various elements. There are AI vendors who sell ‘products’ but of course these are not bespoke and every company buying these will get the same insights and strategy. They are good for some things but not for everything. Ideally each company have their own data science team who can work with the C suite and assist the strategy teams know what is possible and what is not.

6. Ensuring use capability
Many companies spend 90% of resources on building their analytics and only 10% on helping the end user to understand and use them. This is a critical component as no matter how great the model, if the end user does not understand how it helps them do their job better and meet their own job’s strategic objectives, it may not be used. Ensure you spend time socializing the approach within the company and spend time training the users and showing them how this will help them do their job faster and better and get stronger results in the areas that matter to them.