5 Ways Dirty Data Derails Digital Transformation

Submitted by graham on Thu, 01/31/2019 - 16:16

There is broad awareness among CxOs that business transformation facilitated by digital transformation is essential for long term success. However, at the outset of a digital transformation programme it is too easy for the focus to be on how to optimise customer experience and the tools, technologies and processes that can deliver against that vision.

It is not always apparent that the condition of a company's data is a significant determinant of the success of the transformation programme.

A data project should be a high priority, budgeted aspect of a digital transformation programme.

Two large risks of overlooking your data at the outset are:

  1. By the time it is apparent that data quality is hindering the transformation programme there might be no more budget available.
  2. Dirty data can hinder projects in several ways, which can slow progress and degrade outcomes to the point that confidence falls and the chances of project abandonment increase.

 

Here are some of the ways that dirty data derails digital transformation projects:

  1. Dirty data hinders integration of new applications
  2. Dirty data wastes your team's time
  3. Dirty data blocks data driven decision making
  4. Dirty data degrades customer experience
  5. Dirty data can damage your brand

Dirty data hinders integration of new applications

When applications are being integrated it is typical for them to share at least some data that is common to both applications. If your data is dirty and you don't plan to clean it up then effort will be wasted assessing how to configure the new application to accommodate it. This might involve synchronising data that could actually be purged, it could be handling a volume of data that is spurious and it could be a necessity to create complex integration logic to handle deficiencies in data.

Dirty data wastes your team's time

There are several ways this can happen. When you know that your applications and processes result in bad data it is common to assign some manpower to cleaning it up at source, or more often as data is exported and moved between departments, the consuming department performs manual corrections on the data. Of course, the source data is still bad in this case, so the effort will be wasted time and time again.

There is another drain on your team's resource as well though. If you have not identified the cause of data quality problems it is likely that you and possibly your customers are seeing the symptoms, which can manifest themselves as bugs or failures, which can be reported to your IT ops team who have to spend time tracking down the cause and either fixing it applying some workaround.

Dirty data blocks data driven decision making

Data driven decision making is a hot topic and rightly so, it can uncover new opportunities, help manage risks, optimise budgeting etc. But if your data is in bad shape you cannot trust the conclusions that you might draw from analysing it.

The worst case scenario is that you invest in building your data warehouse, you hire the experts to setup your BI reports and dashboards and then you verify what the data is telling you and realise that the insights are junk. Customer valuations are split across duplicate records, non-standardised reference data creates spurious reporting dimensions and your metrics are therefore almost meaningless.

Dirty data degrades customer experience

A digital transformation programme is not simply throwing up a new ecommerce website, offering a mobile app or automating a process or two. Those activities are digital projects that could form part of a digital transformation programme, but a digital transformation programme starts by asking "how does modern technology allow my business to service customers, both internal and external, most effectively".

In that context a large part of customer experience is about tailoring service based on what you know about a customer and their preferences. But if you cannot efficiently access and analyse all of the pertinent data due to data quality problems then you cannot service your customers most effectively regardless of the tools and processes you choose to adopt.

Dirty data can damage your brand

A digital transformation programme will see your business engage with customers in ways that previously you did not. When your data is dirty you risk making ill judged engagements, such as mailing those who are deceased or who have opted out of certain communications. You risk using incorrect data, such as a maiden name or an old address, even when the customer has taken the time to update these records with you. Sometimes you risk looking incompetent if you have so much duplication of data that you bombard people with countless versions of the same communication.

Such errors can result in customers questioning your ability to properly handle their data and your competency to service them.

What is the answer?

The first thing to do is pause.

Don't plough on just because you didn't budget for a data project at the outset, the risks of failure are too great.

Budgets are always tight, so how do you proceed in the face of dirty data when the digital transformation programme is already pushing budgets to the max?

The good news is that if you have a data quality problem that needs solving, that almost certainly means there are clear opportunities for cost reduction across the organisation and that can seed your data project budget. Where the same data is being manually cleansed over and over again, possibly by multiple people across different departments, estimate the yearly cost of this wasted effort and then target a solution at source that can yield a net saving to the business and free up team members to do more productive work elsewhere. 

Look for similar savings the organisation will enjoy by freeing ops teams from the burden of handling issues that arise due to data quality issues.

Next, look for cost savings in integration projects that no longer require complex workarounds to handle dirty data. Over several integrations this adds up.

Finally, prioritise. There is no point delivering the rest of the transformation programme if data quality issues will give it a negative return on investment. Instead, prioritise data quality and those other projects that maximise new revenue generating possibilities and ensure that lower priority tasks have a sound business case.

Digital transformation does not demand that everything that can be digitalised, or automated, is actually digitalised or automated. If the business case is marginal or worse then scratch the project and save the budget. To reiterate, the goal of digital transformation is not to adopt new technologies and automate, they are simply tools that can be used to achieve the real goal, which is servicing your customers most effectively.