All partnering requests
Estimating incremental innovation impact and uncertainty
  • Background
  • What we're looking for
  • What we can offer you
  • Who we are
  • Q&A
Have questions about this request?
Get them answered by the team at Procter & Gamble.
Background

Organizations make significant innovation investment decisions before full consumer and market impacts can be observed. Existing forecasting, market research, and econometric modeling approaches provide valuable signals, but often struggle to isolate the true incremental impact on consumer behavior and business metrics (such as sales, share, or category growth). 

 

The core technical hurdle is establishing a credible counterfactual: what would happen without the innovation compared with what happens because of it. Real-world market outcomes are heavily confounded by commercial interventions—such as pricing, media support, distribution changes, competitive reactions, and broader macroeconomic dynamics. 

 

Currently, teams cannot reliably determine, before launch, the incremental impact of an innovation, the true uncertainty surrounding that impact, or which specific additional evidence would most improve investment decisions. We need analytical approaches that move beyond correlation to provide rigorous causal decision support under real-world uncertainty.

What we're looking for

We seek a validated decision-support method and working proof of concept to anticipate the incremental consumer and business impact of innovations before launch. The approach should distinguish what can be estimated with useful confidence from what remains inherently uncertain, quantify that uncertainty, and identify the evidence most valuable for improving the decision. Where applicable, it should establish a credible without-innovation counterfactual and isolate the innovation’s contribution from commercial and market factors. Reliable prediction is not expected for every innovation.

Speed to evidence is critical: an initial testable proof point should be achievable within approximately 30 days of receiving the required data and inputs. Selected, appropriately masked historical innovation data may be provided to support model development and testing, subject to confidentiality and availability.

Solutions of interest include:
  • Bayesian decision-support systems incorporating uncertainty quantification and VoI analysis
  • Causal inference and causal machine learning models isolating incremental contribution
  • Counterfactual modeling, synthetic controls, and heterogeneous treatment effect estimation
  • Econometric and behavioral simulation models evaluating pre-launch innovation interventions
Our must-have requirements are:
  • Prospective & Calibrated Estimation: Generates upfront prospective estimates of incremental impact of innovation on pre-agreed consumer and/or business outcomes where meaningful estimation is possible; quantifies uncertainty and explicitly identifies where reliable prediction is not feasible.
  • Counterfactual & Causal Logic: Where applicable, establishes a credible without-innovation counterfactual and distinguishes the innovation's incremental contribution from relevant commercial factors and market context.
  • Blinded Historical Validation: Demonstrates the potential to evaluate prospective assessments on blinded historical innovation cases before actual outcomes are disclosed and compares performance and calibration against agreed baselines, where reliable quantitative comparison is possible.
  • Transparent Uncertainty & Learning: Identifies key sources of uncertainties, underlying assumptions, and evidence driving it, with an approach to indicate what additional evidence would most improve confidence in the decision.
  • Rapid Speed to Evidence: Demonstrates the potential to deliver an initial testable proof point within approximately 30 days of receiving the required data and inputs; innovators may propose an alternative timeline with justification. Subsequent development should follow short, evidence-led cycles with clear continue, adapt or stop decisions.
Our nice-to-have's are:
  • Ability to be incorporated into repeatable innovation investment decisions.
  • Explainable outputs and transparent treatment of assumptions and uncertainty (no pure black boxes).
  • Generalizable methodology across varied consumer goods categories, product types, and global markets.
  • Robust performance with heterogeneous, incomplete, noisy, or relatively small commercial datasets.
What's out of scope:
  • Conventional market forecasting or time-series extrapolations that fail to isolate incremental causality.
  • Generic advisory services without a differentiated, testable method, model or analytical capability.
  • Retrospective observational correlation analysis presented as causal proof.
Acceptable technology readiness levels (TRL):
Levels 3-5
What we can offer you
Eligible partnership models:
Co-developmentLicensingSponsored researchPilot or trial engagement
Benefits:
Sponsored Research
Funding of up to €50k may be available for the initial proof of concept, depending on scope, data requirements and partner type. This is indicative and non-binding. Successful feasibility demonstration may lead to follow-on funding.
Data
Selected masked historical innovation data may be provided, subject to confidentiality and availability.
Expertise
Collaboration with relevant P&G innovation, R&D and data/analytics experts, including business context and structured feedback to evaluate the approach against real innovation decisions.
Who we are

At P&G, innovation is what we do best. We love finding solutions to problems. With the heart of a start-up and the resources of a global corporation, we are always looking for ways to reinvent every aspect of our business. As we innovate, we find inspiration in people - their needs, values, desires, and passions. The people of P&G R&D know that collaboration is key to unlocking visionary thinking. Our experts seamlessly work together with our external partners in the pursuit of the new, the next, the impossible. Read more here: https://www.pgconnectdevelop.com/

 Learn more
Reviewers
Q&A with Procter & Gamble
Ask the team at Procter & Gamble any questions you have about this partnering request.
Ask your question
Sign up to access the full partnering request.
View the details of this request and connect directly with corporate R&D teams at Fortune 500 companies.
Submit Proposal
Deadline: January 8
Refine recommendations
Is this request relevant to you?
Eligible partners
Academic researchersConsultantsService providers (CROs/CMOs)Startups
Seeking partners focused on
Artificial IntelligenceCognitive NeuroscienceCollective BehaviorComplex SystemsMachine LearningNetwork OptimizationNetwork StructureSocial SystemsSociobiologyStatistics & Probability
5 more

Halo home
Partner smarter. Move faster.
Get new partnering requests
delivered to your inbox.