Think your firm is data-driven? Think again! Addressing knowledge loss in your organization
Ninety percent of the companies we’ve spoke to at Nomad Data are tracking all their data providers in a spreadsheet. And no, this isn’t just small startups. We polled nearly 100 “data-driven” firms, from global banks, consumer product companies and investment firms managing billions of dollars, and the most popular solution to track all of a firm’s data providers, past data evaluations and data learnings is a spreadsheet. And at a close number two is… that’s right: in people’s heads. In this article we will go over the important role knowledge management plays in extracting value from data and how to make your firm data powered.
The Importance of Data Management For Businesses
The impact data will have on businesses will be as profound as that of the Internet. Businesses that didn’t embrace the Internet, and rather focused on defending antiquated business models, either don’t exist today or are in the throes of their demise. As we enter the age of data, businesses face a similar fork in the road. As Mark Twain is reputed to have said, “History never repeats itself, but it does often rhyme.” Those companies that view data as a nice to have, a bolt on, or a handful of self-serve dashboards will soon find their market share shrinking, their profits eroding and their talent pool evaporating.
Defining the Endpoint: The Essence of Data Strategy
The key tenet of a true data strategy is process. The starting point of a process is ironically to define the ending point. What are we trying to accomplish? This can’t be answered with buzzwords and phrases like “We want to be data driven”, “We want to increase sales using data”, “We want our investment decision process to be improved through data”. These need to be a set of real, tangible goals like “We want to be able to track our main competitor’s store expansion at the permitting stage so we don’t cede market share” or “We need to be able to identify new data which will allow us to reduce the likelihood of underwriting loans to borrowers who will be delinquent”.
The Importance of Process in Data Strategy
With a north star identified you must start filling out the other pieces of the process. These may include addressing new data requests over time, identifying potential data providers, testing data, paperwork management, feature engineering, ETL, integrations and many others. But just setting up these steps isn’t enough: knowledge management is critical to a high functioning data organization. Through each step in your data process things are being learned. These learnings must be captured centrally so that others can learn quickly without repeating data discovery, data evaluations, purchase paperwork and data wrangling.
The Role of Knowledge Management in a Data-driven Organization
To make your data process a durable one, it can’t rely on knowledge in individuals’ heads, which is so often the case. For your data effort to succeed you must ensure that everything learned is retained so that as employees depart or change roles, your organization’s efforts aren’t set back. Almost all firms we spoke with had no process for capturing knowledge. As employees moved on, critical know-how was lost. Newly onboarded employees had little to no context for what had been tried at the company before. They weren’t aware of the firm’s previous interactions with data providers, they had no context of who at a firm touched a particular dataset over time. This creates a doubly negative impact from departures since not only is knowledge lost, but it becomes increasingly challenging for new hires to decipher what’s happened before.
The Need for a Central Repository of Knowledge
So what does knowledge management look like in a highly functioning organization? Let’s walk through the lifecycle of employees in a firm that does employ knowledge management around data:
New employees can immediately see:
- All data providers their firm has a relationship with or has interacted with previously
- All history around sales discussions, even when the data was not purchased. This includes all meeting notes, sales collateral, all documentation, legal and compliance paperwork
- Full visibility into all testing that was done on the datasets during evaluation
- All internal data products, who is using them, for what and how exactly they did it
All requests by employees for data to support an initiative are logged and tagged to the data provider that was used, or even just evaluated, to address it. If external providers responded to the request, what was learned about the providers from their response is also logged. If the data to address the request was internal, how was the data used, what issues arose and how were they resolved.
When a new project is initiated around an existing dataset, employees can see:
- All other employees that are currently using that data and for what purpose
- All learnings from past projects implementing or utilizing the data
- Details on all internal contacts needed to access the data or provide further context on past learnings
As the new employees learn, they will use the knowledge management system to record:
- Every new provider they interact with
- Context around every meeting with the provider
- Content around internal meetings regarding testing of new data, or a new implementation of existing data
- Dataset costs, purchasing, and access details
- Strength and weaknesses of internally created data products and a history of who is working with them
With all this detail cataloged, the loss of the employee comes down to the loss of a particular skillset. When an employee with similar skills is brought on, they can quickly review all context within the firm to avoid repeating past work. They can also quickly build on the learnings of others to move much more quickly.
Drawing Parallels: The Sales Process and Knowledge Management
The idea of implementing knowledge management is not a new one. Imagine a sales process without a CRM? As your salespeople leave, you lose all knowledge of sales leads, previous discussions, deal stage information and more. Given how obvious the need is for knowledge management in data, why is it such a rarity? Many of the people we spoke to admitted that if they were to leave their job, the firm would lose almost all details on their provider relationships and interactions. Even though many did store some of this data in a spreadsheet, almost all admitted that no one at their firm had knowledge of the spreadsheet’s existence. Many felt this led to a significant improvement in job security. While this is certainly true for the individual, it creates an unscalable situation for the employer.
Centralizing Data Relationships With Nomad Data’s Data Relationship Manager
Nomad Data’s Data Relationship Manager helps firms centralize all their data relationships in one place. Discoverability of new and existing data relationships is real-time. Employees can spend their time creating value rather than searching for the context, data and tools that they need to do their jobs.
Data Relationship Management is a must for firms that want to execute a data strategy. It enables firms to get increasingly smart around data over time, avoid past mistakes and prevents duplicative spend. We see this as being one of the fastest growing categories in data over the coming five years.