Companies need to be more focused on proprietary data — data that is unique to a company and can be used to create a sustainable competitive advantage. The need for proprietary data strategies is increasing with new data types and the growth of artificial intelligence (AI). Most commercial AI involves machine learning, and if your company has the same data as everyone else, it will have the same models informing these machines, and thus no competitive advantage. A company’s proprietary data strategy should address the full lifecycle of such data — from what might be done with it, to how to get it, to the ethical considerations that might result from it. Beyond simply appreciating the need for such data, a strategy effort can answer key questions about how proprietary data fits into the strategy and business models of an organization.
What’s the most overlooked piece of your company’s data strategy? If you’re like many companies, it’s probably proprietary data — data that is unique to a company and can be used to create a sustainable competitive advantage. This is not to mean trade secrets and intellectual property (which is often proprietary but seldom really data), but rather, data where the company is the only organization that has it, or it has added enough value to make it a unique business asset. Proprietary data can be big or small, structured or unstructured, raw or refined. What’s important is that it is not easily replicated by another entity. That’s what makes it a powerful means of achieving offensive value from data management.
We and others have been writing and talking about the value of proprietary data for many years. But we still see few organizations with strategies for how to acquire, develop, and leverage it. Most companies focus only on their internal data, which is proprietary in a sense, but may not be a valuable asset unless further developed. If, for example, your internal data sheds light on an issue that other organizations face (payments data for a credit card firm, for example), or if you can combine it with external data in a way that makes it useful to other firms, it could be a proprietary asset.
Sponsored by SAS
Best practices for putting data to work.
The need for proprietary data strategies is increasing with new data types and the growth of artificial intelligence (AI). There are many new types of data emerging across industries — sensor data, mobile data, new types of payment data, and more. Most commercial AI involves machine learning, and if your company has the same data as everyone else, it will end up with the same models informing these machines, too — and thus no competitive advantage. Organizations need to think about their proprietary data strategy and put it into action now.
Some companies and industries are already pointing the way to effective proprietary data strategy. Alphabet’s Waymo and GM’s Cruise Automation, for example, are assiduously gathering maps and sensor data from billions of miles of simulated and on-road driving. Firms focused on medical imaging for AI-assisted radiology or pathology are acquiring or partnering for image data. Companies in media assiduously protect the value of their films, TV series, news, books, magazines, and so forth, and are increasingly distributing those content assets in a variety of formats and channels, many of which are envisioned at the beginning of a content creation project.
Or, look at investment firms. They are increasingly interested in accumulating and analyzing “alternative data,” or nontraditional ways to determine how the economy or particular companies are performing. They might assess the performance of the retail industry, for example, by analyzing satellite photos of store parking lots. Successful hedge funds like Renaissance Technologies have prospered in part because they gathered, curated, integrated, and analyzed datasets like securities pricing data and made it a proprietary asset.
A company’s proprietary data strategy should address the full lifecycle of such data — from what might be done with it, to how to get it, to the ethical considerations that might result from it. As you develop your strategy, consider the following questions:
- For what business purposes would proprietary data be useful? New products, business models, customer relationship enhancements, or something else? How would competitive advantage be achieved and sustained with it?
- What data types would be valuable to the organization?
- How will the organization add value to, curate, and protect internal data of value?
- What types of publicly accessible data might be useful, and how will we add value to it and make it proprietary?
- Who might possess external data that would be useful to us, and how can we ethically obtain it? Consider buying a license, buying the company, scraping it, etc. (But be careful with scraping: the facial recognition software company Clearview AI, for example, has been criticized for scraping facial images from the Internet.)
- How do we firmly establish our claim on proprietary data for which there might be alternative owners, such as customers? Are there legal agreements in place that firmly establish our ability to use the data as we wish?
- How do we monetize the value of proprietary data?
- Does your data architecture make it easy to pull all your proprietary data together?
Keep in mind that proprietary data often has its own “network effects.” Acquiring or integrating more data and curating it effectively creates a more valuable asset, which gets embedded into “data products” and customer offerings. Those products and relationships bring in more data, which can then be added to the proprietary data store. Google Search, for example, gets better and better at divining customers’ search intent as more people search and click on search results. Thus far it has primarily been multi-sided platform businesses that have experienced this virtuous circle, but it can be true for any type of firm.
Of course, as with many of those platform firms, more data can mean more work. Facebook, Airbnb, and Uber have certainly achieved a lot of proprietary data, but their data-intensive business model now requires them to ensure that news isn’t fake, that dwellings are accurately described, and that drivers are safe. Proprietary data strategies must consider both the opportunities and the potential burdens of new types or volumes of data.
Finally, it’s important to ask who should be managing this proprietary data in your organization. Since most companies don’t have anyone responsible for proprietary data, they’ll have to assemble a team to create strategy for it. Chief Data Officers, while typically responsible only for internal data in organizations, may be well prepared to lead the strategy process if they have revenue responsibility and strong relationships with business executives. Other participants might include representatives of the IT, legal, product development, and marketing organizations. If the proprietary data strategy is likely to involve new offerings for sales to customers, the sales function should be involved as well. Many “data products” have foundered because salespeople aren’t comfortable selling them; they often want to throw them in for free in order to help sell more tangible offerings.
As the majority of organizations are rapidly becoming more data-intensive, proprietary data is essential. Beyond simply appreciating the need for such data, a strategy effort can answer key questions about how proprietary data fits into the strategy and business models of an organization. We expect to see many more proprietary data strategies — and successes in applying this resource — in the near future.
Vía Harvard Business Review https://ift.tt/3foip5J