Savvy Companies Exploring the Use of AI to Monetize Own Data
Companies are exploring how to apply AI to monetize data the company generates—through improved sales or reduced costs–to gain a competitive edge.
By John P. Desmond, Editor, AI in Busines
Companies trying to decide how to use AI strategically are being advised to use their data to gain a competitive edge.
The authors of a recent account in Harvard Business Review recommend proceeding in four steps: survey the company’s data and how other companies are generating and using data; look for startups to help jumpstart the data strategy; buy, don’t build; and start building a data moat.
The authors note the continuous progress of AI, with over half of all companies at the end of 2021 having adopted AI in at least one business function, and more companies reporting that AI adoption is contributing to earnings.
“With AI tools, non-tech companies can use the data they already have to improve sales, logistics, and operations in general ways,” state authors Ulrik Stig Hansen and Eric Landau, cofounders of Encord, a London-based company offering data-centric AI automation tools
Some companies are examining historical data on utility use and building maintenance costs to try to save on future expenses. The authors cited the example of Google’s DeepMind AI unit, which took historical data about temperature, power, pump speeds and more that had been collected by thousands of sensors, and used it to train deep neural networks in an effort to reduce data center energy consumption. The result was a series of recommendations, the authors stated, that reduced the energy required to cool the data centers by 40 percent.
“Consider engaging early-stage startups with proof of concept contracts or creating data-sharing agreements with seed-stage startups to understand what innovations are happening,” the authors suggest.
Buying off the shelf tools can save time and money. “Don’t trick yourself into thinking your use case is so specific that it requires a special internal infrastructure,” the authors state.
The data moat, a barrier to entry, can be used to protect data supporting high value-generating activities. “Eventually, this moat might become so large that it’s too wide for other companies to cross it, so the data provides you with a competitive edge,” the authors suggest.
Tesla Has a Data Moat
An example of a company with a data moat is Tesla. With some two million electric vehicles sold, the company leverages data from the owners who use their vehicle’s self-driving software. This included information about accidents, human behavior and more. “Having this real-world data at scale sets Tesla apart from the competition,” the authors state. It also gives the company an asset, if they ever decided to sell its data inventory to other companies.
“So don’t throw away your data. Collect and store it until later down the line when you can use it to meet future business objectives,” the authors advise, adding, “Don’t treat it as a useless byproduct just because it isn’t your primary product right now.”
Lower to the ground than Tesla is the Civic Consulting Alliance nonprofit in Cook County, Illinois, which hired an analytics consulting firm to help build a dashboard that delivers on a racial equity framework. The visualization tool, built with Tableau, allowed the Alliance to identify communities that have experienced disinvestment, and to visualize trends in household income by census tract.
“I couldn’t be happier to witness how our analytics solution has inspired data-driven thinking across the firm,” stated Kirsten Carroll, principal at Civic Consulting, in an account on the website of Analytics8, a data analytics and consulting firm.
Offer Data or Insights as a Product or Service, Analytics8 Suggests
The founders of Analytics8 suggest in a blog post that companies can monetize their data externally in two ways: either by offering data as a service or a product, or by offering insights as a product or service
The authors, Kevin Lobo, managing director of analytics, and Josh Goldner, Google practice director, suggested five best practices to get started with data monetization:
Quantify the value of your data;
Manage your data;
Consider access and security;
Assess your current technology stack;
And get buy-in from your organization.
On the technology stack point, “Ensure that your tech stack can feasibly support the use cases you have in mind,” the authors advise. For example, if you are interested in providing real-time, streaming metrics in your offering, do you have the cloud-based technologies in place to support it? And if you find gaps in your technology stack, are you willing to make the investments needed to fill those gaps?
The right people skills are needed to successfully monetize data using AI projects. A recent account from the MIT Sloan School suggests companies are better equipped when they have enterprise data monetization skills in five areas: data science, data management, data platforms, customer understanding and acceptable data use.
Researchers Barbara Wixom, Ida Someh, and Cynthia Beath from the MIT Center for Information Systems Research looked at an example from Microsoft, to build a new AI program in its real estate and facilities group. With a mission to oversee buildings housing 190,000 workers, the group sought to build AI programs that leverage data to help reduce building facility costs.
The team’s 20 data scientists studied how space was used, and built a model to be used for building and parking garage optimization. The data management team, seeking curated data sets that are accurate and well-understood, identified the data needed to train machine learning models. They acquired, cleaned and validated the data, and also integrated it with external data on weather.
The data platform team created a data lake to store and analyze data, and make it available for sharing, first internally and later to external users for accessing customer services. Internal and external customers were engaged as the models were developed, to get their perspectives and help create products. The AI programs were found to have potential commercial application, so the team worked with customers and Microsoft’s consulting team to develop related products.
Another team worked to establish acceptable data use practices that aim to minimize decision risk, bias and unintended consequences. “When data is being used commercially, the data protection practices need to scale up,” to ensure privacy and security, the authors stated.
Read the source articles and information in Harvard Business Review, in a blog post from Analytics8 and in an account from the MIT Sloan School.
(Write to the editor here.)