In the shift to an increasingly data-driven future, robust portfolio analytics processes are key in enabling organizations to maximize their value by identifying gaps and implications (e.g., risks, upside) of potential portfolio decisions. As pharmaceutical organizations navigate the institutionalization of these processes, they face several challenges that inhibit them from achieving their goals.
Data sources and methodologies for key analyses inputs can differ by business function, resulting in differing decision outcomes. Teams may disagree, for example, on whether to invest in phase 3 trials if they both assume a different probability of technical and regulatory success. Internal portfolio decisions can sometimes be made without considering either the broader market context (such as the presence of competitive assets) or its potential implications on a decision’s feasibility. Inconsistent decision-making can lead to increased compliance and audit risk as larger decisions, and the rationale behind them, are often subject to intense scrutiny.
Key decision-makers can sometimes lack visibility into relevant context that materially influences whether to pursue certain opportunities. For example, a program lead (someone who oversees an asset's development and strategic planning) may pursue the development of one indication over another, not knowing that the passed-upon indication would have been synergistic with the market strategy of another portfolio asset. As needs for analytics increase within the organization, the resources required to service various stakeholders’ requests grow rapidly and, if managed incorrectly, can not only outweigh potential benefits of performing these analytics but also divert resources away from high value-add analyses.
Utilizing automation to improve decision-making processes (e.g., automating data management, data visualizations) can address the pain points mentioned earlier and also empower team members to spend less time quality checking and more time on analyses-informing development decisions.
Standardizing data inputs in analyses via automation can increase decision-making consistency. Automating the underlying methodology and inputs behind key processes and analyses (e.g., standardizing how data inputs feed into the overarching analyses) can help ensure internal decisions are made upon the same foundational assumptions and minimize internal back and forth. Data teams, via automation, can aggregate up-to-date information from external data vendors and enable decision-makers to quickly leverage this data in internal analyses to ensure all stakeholders are correctly factoring in market conditions in decision-making.
Automating data centralization and report generation can improve cross-portfolio transparency. Automated centralized portals provide up-to-date overviews of assets and ongoing developments can help otherwise siloed program leads make more informed and synergistic decisions across the portfolio. Automating analysis generation and refreshes (input, visualization) can enable stakeholders (who would otherwise be forced to rely on internal analytics teams) to more easily leverage portfolio-level data and context when making decisions.
Automation can enable more scalable analytics. Automation of data input and output visualization can not only decrease the lift needed to generate relevant analyses for stakeholders but also reduce the burden on internal analytics teams, as stakeholders will be able to generate reports of interest themselves with minimal effort. This ability to reduce menial tasks will enable the addition of new and exploratory analytics.
Advances in technology that allow for data to be connected across the organization (e.g. data fabric) make it possible to bring together and automate like never before.
There are four important things to consider when determining which decision-making processes to automate.
Ensure internal alignment on the organization’s automation objectives. These are the cost reduction, decision-making improvements, establishing a foundational set of analytics processes and methodology that will persist even after key members leave.
Identify relevant items and align to your objective. Find things that could potentially add value to the organization if automated.
Identify which items make more versus less sense to automate. Processes which are intrinsically very “high human touch,” such as making actual decisions, wouldn't stand to gain much from automation compared to processes which are “lower human touch,” such as standardizing data sets or generating visualizations.
Evaluate short-listed items as either short-term or long-term. There are two dimensions you can use to consider which are short-term and long-term: ease of implementation and value-added. Short-term items are quick wins that are high ease of implementation, high value-add. Longer-term items are high effort, high value-add. Longer-term processes typically involve:
- People-intensive processes (very manual labor-intensive)
- High error rate analyses
- Decision-making support
Automation can be particularly beneficial to portfolio teams in two main areas: automating to improve decision-making consistency and scalability of analytics and automating to improve cross-portfolio visibility.
Automation can help analytics teams improve cross-portfolio visibility when they create a cross-portfolio platform that’s easily and broadly accessible. We recommend that the platform features the provision of an overview of assets and developments, including ongoing clinical trials and their status, indications being developed by asset and treatment combinations being tested by asset. The platform should also generate cross-portfolio impact analyses of various decisions, such as the cannibalization impact of asset and indication development on a contingent asset or portfolio’s value.
As organizations initiate or continue automating in portfolio analytics they should keep the following considerations in mind:
- Automation may take time due to the involvement of many stakeholders. For processes that span many potential teams and stakeholders, automating seemingly “menial” processes, such as updating a timeline or changing the data sources underlying assumptions, will involve a significant amount of socialization and stakeholder discussion as teams must align on the downstream implications of the change (e.g., debating the merits of using one data source assumption over another).
- Actionability is a guiding principle. In addition to removing the manual labor underlying these processes, equally important is the need for automation to imbue an organization's collective expertise into data to make it actionable and efficient. More tactically, automation should proactively serve up data to facilitate stakeholder discussions around what the data means and how to interpret it to ultimately transform raw inputs and outputs into something decision-makers can readily leverage.
- Garner buy-in and develop champions. Change management and garnering stakeholder buy-in, not only in terms of automation’s value to the organization but also in terms of how processes are being automated, will be key to ensuring end-users adopt automated processes and outputs and anticipated benefits of automation are realized.
Automation in portfolio analytics is set to become increasingly important in the near-future, as pharmaceutical organizations seek to become increasingly quantitatively driven and consistent in their portfolio decision-making. Due to the involvement of many stakeholders in this process and the cross-functional impact of these potential changes, it’s essential for pharmaceutical organizations to define where they can “win” early in portfolio automation. This will allow portfolio analytics teams to build belief within the organization of both the value and accessibility of automation.