While your business needs data in order to make good decisions, there are mistakes businesses often make that cause data collection, data storage, and data management to become hindrances to business operations. Here are four big data management mistakes you should avoid making at all costs.
Everyone Has Their Own Systems
Defense contractor Raytheon found in a major cost-cutting review that they had over a dozen different product data management systems used by different company sites and divisions, costing them tens of millions of dollars in unnecessary costs.
The solution to this problem is consolidating all of your data into one system and expand your data center so that it is easier and cheaper to manage while ensuring that everyone uses the same data to make decisions.
New Projects and Legacy Infrastructure
Use of legacy systems often occurs because of inertia. You already own it and your employees know how to use it. Shifting to new software and business processes will disrupt operations for a while and require up-front costs. Some businesses try to see the benefits of big data by trying to migrate to new data storage systems with more advanced data mining tools but attempt to run it on their legacy hardware. This is a mistake. The best way to support distributed computing and massive data repositories is on hardware designed for new data architecture. While you might be able to run a virtual server on your old server, you’ll run into problems if you try to integrate a new and improved server for your “cloud”. Just upgrade the hardware when you migrate the data.
We Need Data, All the Data
There is a bad assumption that you need to gather as much data as possible to make the data-driven decision Peter Drucker advocated at the onset of scientific management. The end result is while some data is helpful, more is better, even if you don’t know what to do with it. This causes organizations to collect data that they have no foreseeable use for, wasting time and effort; they simultaneously spend more money on data storage and data management software. All of this is compounded by the ever more complex data mining tools to find the right data sets that managers hope to analyze to find an answer.
Structure Is Too Limiting
Too many businesses make the mistake of not applying structure to their data. They don’t use strict formats for document names and they try to figure out what objects are. They don’t use descriptive metadata to describe items, so a searcher cannot use obvious key search terms to find data objects. Another problem is failing to cleanse data upon import, so that apostrophe in a title becomes “*&&#@” in the document title. That makes it much harder to search for by its proper name.
Neglecting to classify data beyond document creator or project means people have to wade through more files than necessary, wasting time.
Don’t continue to collect databases and data management systems; combine them to improve efficiencies and reduce costs over the long run. Never set up modern databases, cloud servers or even “data lakes” on legacy hardware. Don’t collect information you don’t need in the hope you can use it later. Apply structure to your data as you upload it and migrate it between systems.