ANNOUNCEMENT: MAHA Launches Darwin, an Evolutionary Reputation Intelligence Platform


Where do you get your data?

We combine deep evolutionary AI (e.g., natural language processing) for sentiment data, with a wide variety of primary sources for adding data to our pipeline that we then combine into a common framework using standard statistical normalization techniques. This makes data from each source comparable across all of the companies and stakeholder groups in our analysis. Examples of our data sources include (but are not limited to): Paid data aggregators/providers, corporate disclosure forms (e.g., 10K reports, SEC reports), Online data reports, Social media and sentiment analyses, First party data provided by you the client. We include the most trusted and widely used corporate ranking metrics that span reputation, dei, sustainability, esg, financial performance, employee engagement, and corporate governance.

I already use some of those data sources, what makes MAHA unique?

We use comprehensive behavioral AND sentiment data assets with an Evolutionary Biology intelligence layer to isolate the key attributes that drive reputation. MAHA can quantify financial risks that are related to reputation. Moreover, MAHA can predict the future influence of attributes on corporate reputation. Our analytics are unbiased and delivered as a SaaS solution for speed and efficiency within the enterprise.

How do you eliminate bias in your analysis?

The MAHA Platform takes a unique approach to analyzing corporate reputation. We make no prior assumptions about the importance of different stakeholder groups or their attributes. Each analysis is done in such a way as to allow our algorithms to determine which attributes differentiate the best competitors from the rest. Our sophisticated form of pattern recognition and is grounded in the Evolutionary framework for studying the process of adaptation.

How are evolutionary biology and reputation management connected?

Evolutionary Biology is the study of adaptation in nature. Our MAHA scientists (PhDs in evolutionary biology) have developed mathematical models that identify the attributes important for survival and extinction, and are now applying those models to corporate data sets. The resulting analytics help corporate changemakers isolate risk and competitive advantage in an increasingly complex and data driven market.

How does your approach differ from pure sentiment analysis?

By combining our extensive quantitative/behavioral data set with more traditional qualitative/sentiment data, we find gaps between perception and reality. Sentiment and perception data that are well matched to quantitative measures of performance indicate that a company is authentic. When there are gaps between perception and reality, our combined data sets reveal novel solutions for course correction.

How often are data updated and what is the timeframe of analyses?

First party data and sentiment data are refreshed constantly (e.g., hourly/daily/weekly). We track employee, consumer, and invester sentiment in real-time to understand how the things you do affect what people think, feel, and say about you. We update our 3rd party data on a monthly basis . Our data time-series spans the last 9 years and will continue to grow into the future.

A typical analysis includes data from the most recent three years, but our methods are highly flexible and can be extended to longer time-frames or narrowed in focus to cover specific months or financial quarters.

Can you incorporate our 1st party data in your analyses?

We have two main ways to integrate your first party data.

  1. We can integrate your existing data streams via APIs on our analytics dashboard
  2. We can work directly with you and pulse relevant stakeholders and gather new data as well as incorporate your own internal measures into our competitive framework

How many companies do you include in your competitive dataset?

Our data set spans more than 58,000 publicly-traded and many private companies as well. The data span the global corporate competitive space and can be broken down by geographic region, by year, by industry, etc…

What is an adaptive landscape?

Adaptive landscapes are visual and mathematical tools used by evolutionary biologists. Picture a rugged landscape of hills and valleys. The hill tops represent areas of high “performance” (Corporate purpose/ financial gains). Valleys on the surface represent areas of poor performance. The three dimensional landscape is built by interactions among the attributes of the individuals. Adaptive landscapes were originally developed to understand why some individuals are better competitors than others and to predict which attributes would evolve to produce the next generation of high performance individuals. We have adapted these models to understand how interactions among corporate stakeholders influence performance and to predict how changes in these various attributes will lead to differences in performance over time.

Are your models predictive?

Evolutionary biologists have long understood that traits combinations that foster competitive adaptation will tend to evolve together over time. The more stable the competitive environment, the more accurate these predicted changes in attribute combinations will be. Our models incorporate aspects of the competitive environment, a time series of data showing which combinations of attributes work well together, and how changes in attribute scores will influence future reputation, corporate purpose, and financial performance.

Will it replace our existing sentiment data/budget?

Ultimately that is up to you!

We can integrate your sentiment data into our framework. The real power of our approach is that we combine many data-sets in a single analytical framework. This also means that you will probably be able to save money by using a single approach rather than cobbling together many a la carte measurement frameworks. MAHA can also be used to ground-truth the efforts that clients are making in-house and in that sense should be seen as complementary.

How do you know if your models are correct?

Our models have been through an extremely rigorous vetting process. Our novel algorithms have been peer reviewed and published in international scientific journals (e.g., Calsbeek 2012 Exploring variation in fitness surfaces over time or space Evolution 4:1126-37; Calsbeek 2009 Empirical comparison of G-matrix test statistics Evolution 63: 2627-2635) and their application to corporate data has been reviewed and published in Harvard Business Review (Argenti, Berman, Calsbeek and Whitehouse: HBR September 2021).

Moreover, we know from client case studies that the prescriptive recommendations from our models improve client reputation and financial performance.