Knowing which organizations perform the best on any particular dimension used to require subjective surveys or painstaking research. Today, the data to answer those questions exists — it’s captured by the software-as-a-service firms whose services companies use to run their businesses. Mainstream software companies are beginning to hold “data mirrors” up to their customers, allowing scoring and benchmarking of their customers’ strategies. We’ve already seen that it’s possible to use external data to evaluate firms on what business models they are employing, and what those business models mean for their valuations.  Those analyses rely on publicly available data sources, but software providers have accumulated growing amounts of private data on almost every aspect of their customers’ technology, operations, people, and strategies. It’s time that these data accumulators begin to share insights back to the creators of all this data, and several firms are beginning to do so.

The most likely software firms to have such data are those that provide transactional capabilities to their customers using a subscription-based SaaS model. SAP, for example, has data on a variety of transactional domains, from customer orders to vacation balances. One of its business units, Fieldglass, provides insights and benchmarks to customers on external workforce management. ADP, a leading provider of payroll capabilities, allows customers to use its DataCloud tool to compare themselves to other firms not only how much employees are paid, but also metrics like their average job tenure, attrition rates, how much they invest in retirement accounts, and at what age they retire. Neustar’s MarketShare software makes it possible for customers to measure the effects of their marketing programs and compare them to other firms. It is even possible to hold up the data mirror to individual technology users. Microsoft, for example, has a program called MyAnalytics that informs customers of its Office productivity software about how much time they spend on various tasks, and the size and strength of their communications networks.

At the same time that data mirrors and scoring have emerged in the corporate world, capital markets are becoming increasingly interested in the analysis of alternative data sets. Active investors such as hedge funds seek to outperform the market and index providers. Stock index giant S&P bought AI-based boutique analytics firm Kensho for just this reason: to get better at using AI to improve investment decisions, and to diversify the kind of data used to make those decisions. (Kensho uses not just raw financial data but data from all sorts of ‘alternative’ sources.)

Software and other companies that develop data mirrors and scores can grow their top and bottom lines with little or no marginal costs by building investable indices that correlate their unique insight and data to investor returns. These can be marketed and monetized via the capital markets in partnership with exchanges and ratings firms.

We believe there are many more opportunities for software companies to adopt this approach—gathering data, relating it to desired outcomes, and returning it to their customers. Salesforce.com, for example, could let its customers assess themselves on their ability to move sales prospects down the pipeline. Workday could provide even more detailed analyses and benchmarking comparisons than ADP or SAP Fieldglass on the workforce. Oracle could let companies know how their average cost of issuing a purchase order, or its average accounts payable levels, compare to other firms.

Allowing a single company to compare itself to other firms on specific attributes is valuable, but an even greater opportunity to create value from data is to assign customers scores based on data and analytics about how they compare to their peers on broad functions or processes, and provide paths to improve their operations and transform their organizations to become digital leaders. The FICO score is an excellent example; the company reduces a consumer’s complex credit history to a single three-digit score that both creditors and debtors can understand. Imagine if all manufacturers had, for example, a supply chain efficiency score, or all companies had a leadership development score. This would provide motivation to leaders to improve their scores, and allow capital markets to make better decisions about the capabilities of companies in which they invest. The appeal of widespread nonfinancial performance measures for assessing companies has been discussed for several decades, but never achieved—despite continuing growth in assets and priorities poorly measured by GAAP accounting.

Of course, there are several steps that software companies need to take in order to make data mirrors possible. Here are some of the key considerations:

• It’s essential to make sure that your company owns this data or has permission to use it. Many software licensing agreements already allow the use of such data for analysis and comparison purposes, but not all do.
• Aggregate the data and use it to allow comparisons to other customers (or at least averages) so that your customers and prospects know where they stand (analogous to the running and cycling leader boards from Strava.com or RunKeeper).
• Software firms may want to display the data only in anonymized form in order to preserve customer confidentiality. Of course, that lowers the value of the data and inhibits the ability to monetize it. If a company is attempting to provide value for investors, anonymity doesn’t work—but the customers of software firms may find it challenging to get customers to agree to be named. In such situations, the use of publicly available external data can be used for scores and rankings.
• Companies may need some capabilities with artificial intelligence to make data mirrors work, particularly if the score or index is being related to financial performance. Machine learning is the ideal technology for creating a set of predictive scores from a collection of data. Other AI technologies can also be used to extract data from transactional systems, or to analyze and quantify textual data.
• Just as companies like Credit.com provide personalized recommendations for how to improve a credit score, companies need prescriptive analytics and recommendations for how to improve their scores on whatever measure being assessed. Machine learning and natural language generation can provide such recommendations—just as they do so now for investing recommendations at companies like Morgan Stanley.

Almost all of the companies we researched, written about, and have advised are at the early stages of this movement, and gaining momentum.  They increasingly appreciate the potential value of ranking and optimizing a customer’s operations and resources with low-touch recommendations. We’ve referred to this phenomenon as corporate robo-advisers, and we see more of them all the time. But software companies are perhaps better-equipped than any other type of firm to offer them.

This data-first approach obviously opens up a variety of issues related to data ownership and privacy, products vs. services, interpretation of data, monetization strategies, and the power of platform monopolies.  But we expect that data-mirror-builders and the scoring systems that they are creating will change numerous industries, processes, and functions. There is so much internal and external data available now it seems inevitable that at least some of it will be used to assess the current and future growth and prosperity of commercial enterprises.

Source: HBR