An Interview Series with Clarity AI Executive Team on the 8 Dimensions of Data Quality
How does Clarity AI ensure its data is of the highest quality?
Clarity AI uses an 8-dimension framework to ensure data is of the highest quality. Those dimensions are coverage, freshness / timeliness, accuracy, data updates, explainability, consistency, point-in-time, and feedback. In this series of interviews with Clarity AI executives, each of these dimensions is explored and explained. Clarity AI’s expert team creates scientific- and evidence-based methodologies that then leverage powerful, scalable artificial intelligence (e.g., machine learning) to collect, clean, analyze and expand existing data sets to power its sustainability tech platform or to integrate directly into users’ existing workflows.
Dimension 7 – Point in time
Clarity AI’s VP of Product, Ángel Agudo, Head of Product Research & Innovation, Patricia Pina, Head of Data Strategy, Juan Diego Martin, and Head of Data Science, Ron Potok, discuss – with Chris Ciompi, Clarity AI’s Chief Marketing Officer – the critical dimension of point in time and its relationship to data quality.
Chris Ciompi: Thanks again, everyone, for taking the time to chat through another dimension of data quality. The next one’s point in time. Ángel, could you please define point in time as it relates to data quality.
Ángel Agudo: Point in time is the capability to see the different values that we provide across time for metrics, organizations, governments… and how they are evolving. Our platform provides updates that happen in real time. For example, news identified about companies might come frequently, which affects the perception and analysis of the company. Point in time information lets you see what are the fluctuations across the company that might be happening due to those real-time events.
Chris Ciompi: You’re talking about fluctuations in real-time and juxtaposing it to point in time, can you explain that again?
Ángel Agudo: Sure. We define point in time as the combination of information that happens every single moment that we update the information. We provide the data that we have about organizations, which might come with very low frequency, such as yearly updates of the information that is reported by the company, or with high frequency that might come from other sources of information that are not necessarily provided by the company, such as news reports coming from NGOs. All of them will motivate changes in the company on individual metrics or in the general perception of the company.
Chris Ciompi: Thank you. Okey, Patricia. Why is point in time important for consumers of sustainability data?
Patricia Pina: Understanding the sustainability journey that companies are taking and where they stand at different points in time is critical for our clients. If we focus, for example, on the need to decarbonize our economy, it is not enough to understand which companies have low carbon emissions today. The real interesting question is about high emitting companies that are starting their decarbonization journey and are gaining momentum. Decarbonizing the real economy requires identifying and supporting these high emitting companies in their transition; it is only through assessing the evolution of the companies’ emission trajectory at different points in time that we can do that. Similarly, point in time allows our clients to identify companies that are not yet decarbonizing so that they can create engagement strategies.
Chris Ciompi: Thank you, Patricia. Juan Diego, how does Clarity AI ensure its data incorporates point in time as a dimension of data quality?
Juan Diego Martín: In order to understand what was available at a different moment in the past, we need to combine two things: data and processes. For the data, everything needs to be attached to a timestamp, a signature, and the moment in time in which this data was published, researched, processed, and exposed to the customer. This combines different frequencies of data, where news can change instantly and market cap can change every day or minute, while information on policies and regulations may take longer to be updated. The other element we take into account is having processes that are robust to things that might affect the ability to rebuild this picture back in the past. For example, if a company changes its name, or mergers happen, we need to be prepared to combine the information and provide an accurate view of what was available at a specific point in time.
Chris Ciompi: Okay, let’s go through that example. You just mentioned a data point is reported, and then a year later that data point is restated. How does our system, how does point in time work in that scenario?
Juan Diego Martín: So, if we look at the historical view as opposed to point in time, we should use the restated information because it provides the most accurate view of the score of that company at that precise moment. But if we look at point in time, what we are trying to understand is what information was available at a given moment, and then we need to use the unrestricted information in order for the information that was originally published to make sense.
Chris Ciompi: Thank you. Okay, Ron, how is point in time at Clarity AI influenced by artificial intelligence?
Ron Potok: Let’s use a clear, specific example here like Scope 3 emissions. Why is Scope 3 special? It’s special because it’s very hard to report. Its reporting is not standardized yet. Many companies do not report all of the Scope 3 fields that are extremely relevant to their industries. So the lack of standardization and the lack of maturity in Scope 3 forces us to do an extra check on it, and what is the extra check? We use our Scope 3 estimate and compare it with the reported data, and we have the ability to override the reported data with our estimate because of the lack of consistency and maturity of Scope 3 reported data. So I think that’s the one and only case where point in time has a heavy influence from our estimation models. Otherwise in reference to point in time, as we discussed in a previous episode of this series, the news program incorporates point in time to change the ESG scores.
Chris Ciompi: Can you say a bit more?
Ron Potok: We allow new controversies to influence the decisions that you’re making as soon as we can. As soon as we find that data, we process it through our platform in a very efficient manner. So you’re not getting the influence of a controversy a month after the controversy has happened. We want you to be able to know that our ESG scores are reflective of the true ESG risk or sustainability associated with the company at the time.
Chris Ciompi: Understood. Thanks, everyone!
