Debunking The ‘Data Quality’ Myth: 3 Ways Cloud BI Helps
Early on, most customers share with us their concerns about data quality. Which, to be sure, is perfectly understandable. Few businesses—and even fewer small, fast-growing businesses—believe that their data is as clean, structured and error-free as it could be. Not many know that Cloud BI can help.
So a familiar dialogue often ensues.
“We like what we hear about Cloud BI, but our data quality is poor,” they tell us. “We have different formats, codings and descriptions across different areas of our business, and data is inconsistent or duplicated across—or even within—our systems.”
Breaking the barrier
The obvious concern: poor data will act as a barrier to Cloud BI, hampering progress and lowering the ROI. Or, as the old saying puts it, “Garbage In, Garbage Out.”
Indeed, some prospective customers go even further. Data quality is such an issue, they believe, that there’s no point investing in a Cloud BI and Self‑Serve Reporting project until those data quality issues are resolved.
But in both cases, our reply is the same. And is one that is based on our experiences of dozens of implementations of Cloud BI, in a wide range of companies.
And it’s this: not only is data quality not a barrier to progress, but implementing a Cloud BI solution will usually actually improve data quality.
What we find—based on dozens of implementations—is that Cloud BI projects work to improve data quality in three distinct ways.
- First, the simple fact of having a Business Intelligence solution makes data errors more visible. Having all your data in one place, and being able to analyse it and drill‑down into causal factors, makes it easy to spot data quality problems and then fix them. In short, quality issues that were previously buried, now stick out.
- Second, Business Intelligence’s inbuilt data reclassification tools can quickly help to improve the quality of data, using rules. During Extract, Transform and Load (ETL) operations, for instance, it’s easy to replace varying codings and descriptions with a single replacement code or descriptor.
- Third, there is almost always enough structure in the underlying data to get started. In short, ERP and core accounting systems invariably require a level of structure to work, and this can be used as a starting point. So while you may not have the perfect product or customer grouping on Day 1, you’ll still be able to derive meaningful results—and, what’s more, will now know how that data must be restructured in order to deliver even more meaningful insights.
Making a start
That said, there’s no reason to be complacent. If you think that you’re likely to be considering a Cloud BI and Self‑Serve Reporting project any time soon, then work undertaken today will undoubtedly speed the resulting time-to-benefit tomorrow.
Here, for instance, are three quick ways to productively prepare the ground for Cloud BI and Self-Serve Reporting, while also raising data quality.
- Dimensions and attributes. Think about how you might want to ‘slice and dice’ your Business Intelligence reports—by Customer, Product, Date, Territory/Country, Sales Representative, or Supplier, for instance. Are those dimensions and attributes consistently reflected in your data right now? If not, make a start.
- Product and item coding. A Product dimension, for instance, will usually need a consistent Product Code/SKU and Name or Description. Again, make a start now. But also consider other information that you might want to identify—Product Group, Category, Unit of Measure or Colour. And don’t forget purchased items, either—thereby preparing the ground for improved procurement analytics.
- Consistent customer descriptions. Get a list of all of your customers, define some key groupings (Sector, End‑user type, Size), send this list to your sales team and get them to correctly and consistently categorise each customer. There’s no need to add the data to transactions today—but you’ll save time when implementing your Business Intelligence project later.
Find out how Cloud BI can assist with any data quality issues you have experienced in the past, download our free E:book below.