What’s the Best Data Strategy for Enterprises: Build, buy, partner or acquire?
Every large organization is investing heavily in building data solutions and tools. They are building data solutions from scratch when they could be taking advantage of readily available tools and solutions. Many organizations are re-inventing the wheel and wasting resources.
Technology is rapidly transforming every industry. Large enterprises are clear that Data will be the most critical element in their growth strategy. Yet many organizations lack clear business goals on how to utilize the solutions they are developing. Every large organization is investing heavily in building data solutions and tools. They are building data solutions from scratch when they could be taking advantage of readily available tools and solutions. Many organizations are re-inventing the wheel and wasting resources.
Here are those fundamental problems.
- For large companies, most of their data is still in silos and underutilized.
- Data compliance laws are making the data strategy transformations slow because of the prioritizing data security, privacy, integrity, quality, regulatory compliance, and governance, etc comes with its own dead weight.
- There is a real talent shortage problem for enterprises.
Adopting build everything from scratch seems to be the wrong choice for organizations, and they are not taking smart decisions on how to go about solving their data problems.
Big Data and AI Executive Survey 2019 by New Vantage Partners shares some alarming insights. The survey participants comprised c-level technology and business executives representing huge corporations such as American Express, Ford Motor, General Electric, AIG, UBS, Cigna, Bloomberg, and Johnson & Johnson.
Here are the insights from the Survey, and it is frightening:
- Organizations are ramping up investment in Big Data and AI to accelerate business agility.
- Large companies are struggling with data-driven business transformation.
- Cultural challenges remain the biggest challenge to business adoption of big data.
- The Chief Data Officer role is maturing but still ill-defined; CDO's may be ill-equipped.
- Most large enterprises are still not data-driven, and won't be anytime soon.
What strategic Data-driven business transformation options are there?
There are challenges in becoming a data-driven company - but that does not always mean more money and more people can solve them. The solution is to define goals, prioritize them and find smart ways to make it work. When it comes to data - An implementation strategy comes down to four options and a few of their hybrids.
Build from scratch
Building from scratch means designing and developing the core solution using in-house capabilities. For instance, a manufacturer like Boeing, who builds aircraft can decide to make a flight sensor data capturing tool in-house. They can build the data capturing tool better than anyone else because they are aircraft experts. However, for analyzing that data - they can use existing tools such as Looker or thoughtspot.
Customize and open source project
Customizing a great open source project might be an excellent way to solve data problems than building something from scratch. If a retailer wants to visualize the sales data - they can customize an open-source solution like D3 JS and get the job done without developing a tool from scratch and also not paying a huge subscription.
Some large organizations are still hesitant in adapting to open source technologies. This is the time to change your mind, or you'll miss out on a lot.
Buy/license from a vendor.
Licencing software/data to solve a problem is a common thing among enterprises. If you want to extract data from websites, you can buy a license to a tool like import.io or get ready to consume data from a vendor like Datahut.
Another example is if a retailer wants to crunch the numbers to find patterns - they can license a tool like sisense or thoughtspot rather than building it themselves. Since SaaS /DaaS became mainstream - many companies are adopting this licensing model.
Co-developing data products is gaining momentum and this is a strategy big companies can try out. Instead of developing all the data and AI tools in-house - big enterprises can find small companies that are disrupting an industry, partner with them to build a solution jointly. Co-Innovation Network™ (COIN™) by TCS is a great initiative in that front.
Companies like Salesforce are already co-developing software products with partners to deliver tailor-made solutions to different industries. The A I' economy is enabling companies to do it faster than ever.
A team of Navy SEALS can get things done which a platoon of soldiers cannot. Consider laser focussed startups as specially skilled and trained SEAL team.
Many large organizations don't have the expertise required to develop an effective data solution in-house. Sometimes acquiring a small startup that has the technical capability and business acumen is a great way to solve the data problem. An excellent acquisition can reduce the time to market by months or even years.
The acquisition has been a top big data & AI growth strategy for giants like Microsoft, Google, Salesforce, etc.
An excellent example of acquisition as a growth strategy is the acquisition of Quandl. Last year, Nasdaq acquired Quandl, a provider of alternative and core financial data. This allowed Nasdaq to position themselves as a technology and insights enabler for investors. That is a quantum leap for anyone who knew NASDAQ as just a stock exchange
How to move forward
For the organization, the value should be prioritized above control. Using available solutions in the market and capabilities developed by others will often achieve the desired business results faster and more cost-efficiently.
To get similar outcomes from building something inhouse, enterprises will have to invest more resources ( time & money) in developing it. It makes no sense unless the product is a strategic differentiator. Too many data initiatives are being scrapped by enterprises because of the bad choices they made on how to solve the problem.
That's why a smart approach would be to invest in three ways:
- Build only the core functionalities in-house to create long-term value and differentiation and integrate with other readily available tools to get to market faster.
- Co-develop non-competing data solutions by partnering with other organizations ( even competitors ). If there is a need for a data solution that can detect fraud - even competing banks can partner to build it. It adds value to everyone.
- Acquire - If your organization is not hitting a specific data solutions goal, which is important for the company - acquisition of small but capable data startups might be a way to solve the problem. We're seeing a lot of $5 million to $30 million acquisitions in the data solutions space which is an evidence of this.
Data and numbers always speak the truth. Organizations will face negative consequences of not being data-driven if they are not rethinking their priorities and methods to solve data problems.
Bio: Tony Paul (@tonypaul_hb) is the Co-founder of Datahut. He has been helping companies solving their data extraction problems for the last six years. He is a chess lover and an ardent fan of the lean startup. He writes about web scraping, online retail, sales, marketing tips, etc.
- Data Science Strategy Safari: Aligning Data Science Strategy to Org Strategy
- Data-driven to Model-driven: The Strategic Shift Being Made by Leading Organizations
- Why Data Scientists Must Know About Change Management