Columbus, OH— The ‘Lean Startup’ method is in… and not just for Silicon Valley. Everywhere you look, companies big and small are embracing the ideas of customer developmentminimum viable products (MVPs) and pivoting as tools for innovation.

Even companies like Experian have developed Experian DataLabs, a part research lab/accelerator/skunkworks that meets with customers, builds MVPs and tests them in market. Once they find a customer problem that looks promising, they often test an MVP within 90 days.

The lean startup process is cyclical and relies on the build-measure-learnprocess to repeatedly launch MVPs and learn how customers react.

An MVP is used to test a hypothesis and ask a question. It is the simplest possible offering you can provide customers to test an assumption.

It isn’t about making a cheaper product… it is about smart learning.

This process lets innovators isolate different factors of their offerings they want to test to see if those are scalable enough to invest in. The process is designed to minimize risk and optimize investment.

Is there a way for pharmaceutical marketers to leverage the build-measure-learn cycle to get evidence-based feedback on the factors that drive success? Can we get leaner in our approach?

The Shift Toward Accountability in Pharma

Major disruptors in the pharmaceutical industry have changed its culture from big blockbuster budgets toward a focus on quantifying value and measuring impact. There is a strong desire to innovate in the face of these challenges… but the tolerance for risk is low.

Unfortunately in pharmaceutical communications, the barriers to experimentation are high. With tons of approvals and regulatory hurdles, strategic experimentation is often not feasible or cost effective.

Even if it was feasible, the amount and diversity of stakeholders affecting the success of a campaign is very complex. Leveraging MVP-based experiments might not give us all the information we need to know if an offering would be optimal. This complexity creates potential blind spots.

So if lean startup methods aren’t feasible, what can we use instead?

What tools do we have at our disposal that allows us to simulate offerings, test various assumptions and maximize results?

Enter Predictive Analytics

The key to lean startup methods is the idea of simulation— utilizing real customers to simulate reactions on how our offering would perform at scale. The key is to maximize knowledge about the optimal offering without having to spend a full budget launching it. These methods can be seen as evidence-based planning tools.

There has been a rise of machine learning and increased availability of large datasets on nearly every metric affecting patients/physicians. This has paved the way for predictive analytics to become a tool that could fit the bill.

Predictive analytics is a method of leveraging information about the past in order to build a simulated model of a given market. Once we have a valid model, we can run experiments to predict what offerings will have optimal impact in the marketplace even in a changing business environment.

Once you invest in a predictive engine and create a model that can accurately predict future outcomes, the ability to experiment is endless. The answers you get help inform your strategy and planning:

  • What marketing mix should we invest in to meet our launch target?
  • How should we respond to a new competitive entrant?
  • What result should we expect from this new tactic or channel?

Why This Matters

In a world of shrinking budgets, accountability for results and pressure for innovation, leveraging predictive analytics as a tool to help with planning and strategy development may become pharmaceutical marketers’ secret weapon. Just like how lean startup approaches helped entrepreneurs increase their odds of success, predictive planning will help marketers do more with less.

The questions then become:

  • How do we incorporate predictive analytics into our current strategic planning process?
  • Will we listen to the model and drastically alter course if we find something surprising?

Although in its early stages, the technology, data and methodologies for this type of ‘lean’ experimentation already exist. It’s just a matter of how fast the industry will adopt it.

About the Author:

Zach Friedman