5 reasons why data-driven innovations fail

Up to 90 percent of consumer-facing innovations fail in the marketplace with a slightly lower number for B2B products and services. Even Apple is not immune. The Apple Newton, the Apple xServe (anyone remember its rack-based server offering?) or its attempt at colonizing the early WWW with its eWorld online service all show innovation to be a difficult and costly business.

Causes for data-driven innovation failure can vary from a bad product-market fit at launch to an inability to evolve products and services to meet changing technology and market needs. Throughout the 1990s, Yahoo! was a market leader via its directory-based system for navigating a relatively small web of information. As the WWW grew exponentially, this system no longer worked for users and a new way of efficiently searching for those relevant snippets of information was needed. Directory systems may have worked in an age of relative information scarcity but in an age of information abundance, discovery was what users wanted. Google capitalized on this with its PageRank algorithm and is now a $800 billion company. Yahoo! was sold to Verizon in 2016 for $5 billion, down from a peak market capitalization of $120 billion in 1999.

Data capture and exploitation have been behind the rise of many of the largest tech giants over the previous decade. As the amount of data generated by online activities continues to rise, thousands of startups and established companies are building new products and services to exploit this valuable resource. Most will not succeed but understanding these five points of failure will help reduce the risks for anyone embarking on this difficult but potentially very profitable journey.

1. Poor product-market fit

While this may seem obvious, too many companies still develop and launch digital services without really creating something that solves a problem or addresses a real need. As Andy Rachleff, CEO of Wealthfront, recently said, “If you address a market that really wants your product — if the dogs are eating the dog food — then you can screw up almost everything in the company and you will succeed. Conversely, if you’re really good at execution but the dogs don’t want to eat the dog food, you have no chance of winning.”

Simply having terabytes of data to play with is not a sufficiently good reason to launch a product onto the market. Google discovered this with their Google Plus social networking initiative launched in 2011 and shuttered in October 2018. Despite having access to the online habits of billions of users and spending hundreds of millions of dollars developing the service, it never gained a foothold in the social networking space because there was no compelling reason for users to switch from existing networks such as Facebook, Twitter and Instagram.

2. Overreliance on third-party data

The ease with which data can be stored, replicated and shared makes it an attractive commodity to build services from. However, this also creates potential points of weakness for companies. Public bodies around the world are increasingly opening up their data assets for commercial re-use. Data.gov in the U.S. and Data.gov.uk in the UK are treasure troves of unique, extensive and free data and many companies are building businesses off the back of them. However, the obvious danger of this is that anyone else can do the same. The trick is to add enough value to the data and build strong customer relationships to create defensible walls to keep the competition out or at least make it more difficult for them to tempt your users away. In the UK, digital startup Tussell is doing a good job of executing this strategy with its live database of government tenders and contracts. While much of the core of Tussell’s offering draws on publicly available information, it has created a value-added offering around this data which provides the insights clients need when bidding for public sector contracts.

3. Building a product, not an ecosystem

Whether you call it a platform or an ecosystem, the point is to build an environment for your offering and its users that is more than the sum of its parts. This often involves encouraging third parties to join in and contribute and, perhaps, pooling data from customers to create a resource that creates value and makes it difficult for competitors’ products to gain traction. Despite initial resistance from Steve Jobs to the idea of an app store open to third party software developers, much of the iPhone’s success is due to the million plus applications that users can download. In the IoT data space, Terbine is building something similar for companies to upload and sell data. This global data exchange, should it succeed, could unlock value from millions of disparate streams of data and be a powerful stimulus for innovation.

4. Treating data as a physical asset

We know that data is not a corporate asset in the same class as buildings and equipment but too many managers lack awareness of the economic characteristics of data and how this plays out in the marketplace. Depending how it was produced or acquired, data may have very high fixed costs but next to zero marginal costs in terms of replication and distribution. The digitization of data and rise of broadband networks has accelerated this trend as Blockbuster and the music industry have discovered to their cost.

Another unique characteristic of data that separates it from physical assets is that once it has been given or sold to a customer it can still generate value for the seller. Google and Facebook have built their empires on top of this factor. By analyzing which search results we click on or which items we read in our newsfeed the two companies are able to improve their services and offer better result for consumers and more relevant targets for advertisers. Use those feedback loops of data to see how your products and services can be improved as well as to identify new revenue streams.

5. Not exploiting network effects

Often referred to as Metcalfe’s law, this states that the value of a network is proportional to the square of the number of users on the network. This is a factor of the number of possible connections which can be made by network users and means that a virtuous circle can be created by companies which are able to gain initial traction in the marketplace with their offerings. While this can make it difficult for new entrants to take market share from successful incumbents (see the example of Google Plus above), it also offers an avenue for startups to find a new niche and build out a customer base from that.

This is particularly true of two-sided marketplaces where value quickly accrues to a company which can build a solid base of buyers and sellers on a single platform. Salesforce has been a leader in this respect with its AppExchange offering thousands of third-party apps that integrate with its CRM offering. While some other CRM vendors offer similar marketplaces, Salesforce has used network effects to create a compelling offering for users and app developers. A large network of users makes it commercially attractive for software builders to invest in developing applications for AppExchange and this, in turn, creates more value for users.

As economies and markets adapt to the rapidly emerging world of data-driven businesses it is vital old ways of thinking about product development and user needs adapt too. This is particularly true for firms adding data capture and analysis functionality to physical product lines. With new data-driven services come new rules of engagement.