Across the world, and here in SA, companies are coming to understand the consequences of poor data quality. Its effects ripple throughout the business, undermining strategies, crippling meaningful business analysis and harming customer-oriented initiatives. Data, once the preserve of IS and its attendant techies, is finally being acknowledged as a critical business issue which belongs foursquare in the boardroom.
Get it right, and data quality becomes a business liberator and enhancer; get it wrong, and the costs escalate exponentially.
Julian Field, GM, Ascential Software SA
Certainly, this is a view supported by Gartner, which asserts that getting data quality right has become Priority Number One for companies looking to embark on business improvement and analysis and customer relationship management (CRM) objectives.
The reasons are very simple:
- . The age-old principle of "garbage in, garbage out": If your data foundations are flawed, your subsequent decisions and actions taken against this data will be flawed.
- . The absolute premise and promise of CRM depends on a unified customer view, so an organisation can have as detailed and unified a view of its customers as they have of it. This can drive a learning relationship, permit cross- and up-selling, and prevent the embarrassment that arises from marketing goods to customers who already have them.
- . The pace and scale of change, planned and unplanned, today is such that a data foundation needs to reflect and support it.
Get it right, and data quality becomes a business liberator and enhancer; get it wrong, and the costs escalate exponentially. I have enumerated the seven cascading consequences of poor data quality. The cost to the business grows dramatically with each successive consequence, both in terms of direct and opportunity cost.
To grasp the concept best, think of seven concentric rings, with a single point at its hub. This single point is 100% pure data. Each subsequent concentric ring, like ripples in a pond, stretches outward because the issue of data quality has not been addressed properly. Let`s give each of the rings a name and analyse its cause and effects.
1. Rolling Your Own: To integrate and properly manage data from various sources, a company may choose to write its own scripts and routines. This is all good and well, as long as the data universe is easily identified and predictable ... but it seldom is. For in IT and business today, change is the name of the game.
2. The Third Force: Seldom will a company`s in-house data be sufficient for detailed and meaningful analysis and correct decisions to be made. For instance, you may need to incorporate spatial data, meteorological data or a Reuters feed for contextualised business analysis. These are external events and triggers; by trying to incorporate them with your own scripts and routines, you could make it difficult to adapt to changing influences.
3. The Change Within: No business stands still. Companies must respond to business change with, perhaps, the creation of a new line of business or a new division. In such a case data routines and scripts need to be rewritten. But given the skills crisis and the staff churn which characterise the industry, the developer, or group of developers who first wrote the data routines may have moved on. If documentation was not perfect, there could be significant challenges in terms of integrating the new or changed line of business into the existing business. It`s not impossible, but it is tough, time-consuming and expensive.
4. Back-Office Focus: All of this is manageable, but now the business itself might change; and many businesses will, as we have seen! The back-office system might change. The business might be acquired. It may downsize, rightsize or move from one ERP or bespoke system to another, possibly a de facto standard. It may now be necessary to feed and accept data from a supporting application such as Siebel. The business may wish to embrace ERP II. Again, any data quality scripts would have to support this.
5. The Sea Change: Still, this is relatively controllable, given the set constraints of any business. What happens if the market itself undergoes tumultuous change? One of the biggest business changes happening today is e-business, which embraces aspects such as e-marketplaces, extended supply chains and the drive towards collaborative commerce. If your business had to adapt to this sea change, how would you modify your data routines accordingly to support the iterative new data sources? Surely not manually? That`s leaving too much to chance.
6. Decisions, Decisions: By now senior management is making significant decisions against data that has fed through these concentric circles. It could be that these decisions are being made against the context of incorrect, incomplete or out-of-date data. At this level the opportunity cost and direct cost are incalculable because of the extended knock-on and ripple effect of poor decisions. The value chain is collapsing in on itself.
7. The Customer Crisis: The outer circle is the one where the greatest impact of poor data quality is felt the most. This is the one that is least controllable and where the greatest impact is to be felt, both in terms of opportunity cost and direct cost. Ask any of the banks how they interact with their full range of customers and they`ll tell you they can`t. They don`t know which of their customers are buying which products and which ones are candidates for new products, although they`ve invested billions in back-end systems and are awash in data. Accordingly, they are spending millions on customer profiling systems, typically with third-party bureaus, so they can gain a unified view of their customers. In the meantime, nimbler competitors that do not have the legacy ball-and-chain of the large banks are "cherry-picking" the most desirable customers. Then there is the direct cost of marketing the wrong products to the wrong customers: to people who cannot possibly respond, or to people who already have the products. There is the direct cost of alienating top customers through providing them with a fragmented experience ... of making long-standing customers fill in forms every time they want a new account ... the list goes on. It is at this level, where customer churn and defection occurs, that the true cost of poor data quality is to be felt at its most extreme.
I`ve tried in a cursory manner to highlight some of the problems that occur when data quality is not accorded its full gravity in an organisation. There are really only two options when it comes to the taxing and vexing business of data quality: get it right up-front, and integrate data quality at the deepest possible level; or reverse-engineer it into your business. Both options are beyond the scope of manual scripting and require a high degree of automation if they are to succeed in the long-term.
(Please note that Industry Insight pieces reflect the view of the author only. For further stories and opinions on this subject, please visit ITWeb`s related sections.)
Share