Estimation an Innovation brand potential based on Bass Model revisited with AI approach

Nafees A

Nafees A

Business Scientist

Guibert T

Guibert T


Given the huge competition, before launching a new product, each company should estimate precisely the potential before engaging cost of development and cost of marketing.

Many CPG organisations define innovation as a key driver of future growth as the industry is challenged there are there. Since the mid-2000s, there has been no real growth in value in the CPG market.

The unexpected grow

The consumer-goods industry thrived for decades with a winning model: build mass brand equity, grow with grocers as they expand, and drive operating efficiencies. The build-scale model no longer works for a variety of reasons and in large part due to globalization that even securities have gone down. From 10% per year between 2000-2009 to 3.2% per year on average since 2010 stated in the MGI1 report. In addition to that all the trends observed in the industry have been accelerated by the COVID-19 crisis. Large brands, particularly, are faltering and are often struggling to generate brand love among younger consumers. Millennials are four times more likely to avoid buying products from big-name food brands than their parents are. Further, e-marketplace giants are generating 65 percent of growth by the top 150 retailers, putting grocers under pressure, and making them more difficult trading partners.

The explosion of Private labels

The explosion of private labels (which drives roughly 30-40% of CPG growth) has shown for the past decade that innovation success is very linked to consumer perception and supportive marketing strategy surfing on Bio, Inclusive food, Nutrition, Affordability…

The critical question(s) around innovations

Launching innovation nowadays require discussing 5 main interlinked working areas:

1. The role of innovation in renewing the consumer base: What is the critical role of innovation in the portfolio as complementary or growth relay brand? All Innovations will be sustainable if the innovator manage to move and not being sensitive to large variations in consumption patterns. It is therefore not necessarily always a cost-benefit analysis. The right KPI to follow change the lenses of success.

2.The diffusion model: How to diffuse an innovation in the context of the digital acceleration and the reshape of retail industry recontextualizing the negotiations and the places on shelf? Building on the paradigm of the more an innovation will be seen to the relevant tribes, the more it will be bought.

3.The trade-off: What arbitrage to be made in between core vs. innovation? This arbitrage shapes the investment and support strategy looking 3-5 years ahead leading to the critical question of …

4.The threshold: What level of investment is required to emerge vs. market conditions and how to monitor the gradual ramp up and decide whether to stop or continue

5.The tactics: Which touchpoints to leverage to drive awareness and purchase among my tribes? Arbitrage and trade-off between ATL and BTL tactics.

Our approach

Fostering the innovation launch will cover the 5 critical interlinked areas with the focus on 3 main questions: diffusion model, trade-off, and thresholds.

At Pivot, our AI team developed a proprietary methodology to evaluate the three first steps required to developed a data driven innovation product launch.

A.Model the innovation diffusion:
What can you expect in term of diffusion vs. KPI selected?

This analysis requires to understand the key drivers and fundamentals of the innovation (channels, TDP, price, in-store execution, A&P support…) and to create the V0 of the innovation business plan, projecting the key drivers’ trends to 3-5 years.


Our diffusion model is based on Bass diffusion model applicable in technology diffusion that we are adapting to FMCG business based on a detailed research on multiple database to find innovation twins in the past leveraging a KNN approach.


Diffusion is the process through which an innovation is communicated through certain channels over a time amongst the members of the target audience. This process can happen through word of mouth or other marketing activities.

The customers in the diffusion process can be broadly categorized into 2 following segments based on the nature of their diffusion.

  • Innovators are the initial set of customers who are likely to adopt the product based on the marketing efforts of the company. They are risk-takers and want to try out new things.
  • Imitators are the set of customers who bank on innovators and others who have tried this product.

Fig: Number of new adopter’s vs time

Bass model is a diffusion model that forecasts the first purchase of the new product for which no other alternatives exist.

It answers the question that « How many customers will eventually adopt the new product over the period of time? »

It is backed by 2 main parameters:

  • Coefficient of Innovation / external influence (p): represents the behavior of the innovators to an innovation.
  • Coefficient of Imitation / internal influence (q): represents the behavior of the imitators to an innovation.

In this data-driven analysis, we will list and identify all the possible drivers of innovation diffusion ((channels, TDP, price, in-store execution, A&P support, category growth, tribes evolution…) and pick the ones the most linked to our target KPI (penetration, NS…) to build our diffusion model.

B.Analyze the trade-offs and the thresholds of an innovation to determine the optimal level of investment and appropriate business plan


When innovations to be launched have been approved, a successful launch in market need to focus on two critical interlinked questions (before choosing the right tactics).


To conduct this analysis, we are using a data driven approach to quantify, based on the diffusion the right level of support required to sustain the innovation launch and the arbitrage to be made in the core.


This implies 5 working steps


At the end of the project, we’ll be able to answer 3 main questions

  1. What is the potential diffusion of my innovation in each market context (and what are its key components) through a bass model (and a simple calculator)?
  2. How much could I remove to the core without jeopardizing the change to grow?
  3. What is the ideal A&P support and business plan to kickstart the innovation and have payback (KPI to be determined) in 3-5 years



By using our approach, we manage to create an innovation tool anchored in the innovation process and the algorithms are improving with the usage reinforcing the capacity of Innovation leader to seize and size the opportunities.