We all know bad data is bad. But just how costly is it?
Consider: IBM estimates that poor quality data costs the US economy $3.1 trillion annually— a little more than the GDP of France.
Even after inflation, that’s a lot of money. The reason is the inefficiency and poor analysis that bad data creates: employees correcting flawed data, analysts cleaning data, IT integrating incompatible data, and executives making the wrong decisions because they don’t have the data they need.
An article in Harvard Business Review estimated that highly paid “knowledge workers” waste 50%-60% of their time hunting for data, correcting errors and cleaning data. Those percentages are consistent with what we see on a regular basis in our industry of telecom expense reduction.
Better data means better analysis, and a faster path to telecom expense reduction. But if the data is trapped in PDFs, emails, paper invoices and spreadsheets, the cost of doing the expense analysis can outweigh the potential savings.
What are the keys to delivering better telecom expense reduction data and getting better analysis?
Automating Data Collection. Being able to collect data in a cost-effective manner is the first step to “good data”. Telecom expense data can come in many formats: Electronic Data Interchange (EDI), PDF, paper, via the web, etc. And many enterprises have to deal with all of these formats each month. This makes it especially important that enterprises develop an automated process for collecting and centralizing all expense data, so that the next step in the process – translation and standardization – can be performed.
Translation and Standardization. Once all expense data is being efficiently collected and consolidated, it’s ready to be translated and standardized. This step is hugely important. It is where data in PDFs, spreadsheets, EDI, etc. are put into a common digital format for analysis, and where data is mapped to a common set of field names. This is where, for example, charges from different service providers for identical but differently branded services are made “apples to apples”. In practice, making this happen requires specialized software-based tools, such as “telecom invoice readers” that have been developed by Cloud Age Solutions to standardize telecom invoice data from multiple telecom carriers.
Deep expertise in telecom products and services. Trying to “clean” telecom expense data without a deep understanding of the actual technology involved is akin to relying solely on a “spell check” program to review a legal contract. You can easily find trivial errors and miss issues that can cost your business a fortune. Likewise, relying on manual processes for translation and standardization can lead to data entry mistakes, incorrect data mapping, re-work, and ultimately bad decisions based on bad data.
Conclusion: It’s Work – But There’s A Big Pay-Off
So, you decide to focus on good data, what’s the payoff? Is it worth the effort to transform bad data into good? Good data is what I like to call a “jacks or better” situation. You need automation and a strong foundation of telecom knowledge to make the “better data, better analysis” journey. So, at a minimum you need better data just to make informed business and technology decisions today. Where this effort accelerates the payoff is the ability to fully leverage technologies such as Artificial Intelligence (AI) and Machine Learning (ML), which rely on quality data inputs. Companies can then use AI and ML to enhance existing data. You can start to leverage technology to turn data into information into insights. Otherwise, you are just adding to that $3 trillion price tag.
Like this blog? Check out some other recent Cloud Age Solution posts: