4 biggest data migration mistakes
Managing your data effectively is a key component of business success – particularly as a good database can reduce your costs. As your business grows and technologies improve, in time you may need to migrate your data. But fail to do this right and the impact to your business could be huge.
Data migration – the process of moving data from one system, storage device or format to another – can be incredibly tricky, as telecoms giant Vodafone can attest.
In the final quarter of last year, Vodafone customers’ complaints to Ofcom more than doubled compared to the same period in 2014. Vodafone put this customer dissatisfaction down to problems with its recent data migration programme, transferring customers to a new billing system.
Undertaken as part of a £2 billion investment programme to enhance Vodafone’s services and network, the “operational challenges” created by the shift to a new billing system contributed to a 3.2% decline in its UK revenue in Q1 2016.
It’s clear that, even for the most sophisticated companies, data migration can still be challenging. So how can you avoid the pitfalls of an unsatisfactory data migration? Here are four common mistakes to watch out for.
Failing to standardise your data
Before you can migrate your data successfully, you need to assess the data you have. That means evaluating what’s still relevant to your business and what can be removed. After all, there’s no point wasting time and money to migrate data that holds no future value to your business.
Once you’ve streamlined your data, you also need to standardise it. The reason you’re migrating your data may be because, through organic growth, mergers or joint ventures, your company uses more than one data system and you want to centralise your databases. Data inconsistency is a common issue when it comes to data migration failures. So make sure to standardise your data to ensure you don’t introduce problems or inconsistencies into your new system.
Carrying over inaccurate information
As with data inconsistencies, it’s important to cleanse your data to remove inaccurate information, which can undermine the efficacy of your new database. Data cleansing (also called ‘data cleaning’ or ‘data scrubbing’) ensures your data sets are consistent throughout by identifying incomplete or incorrect information and replacing, amending or deleting it.
Depending on the nature and storage of your data, there are different ways you can enact a cleanse – from simple cleaning mechanisms on Excel through to hiring companies with tailored software solutions.
Poor system design
It’s not enough to standardise and cleanse your data. Your new system needs to be designed around your business’ needs, both now and in the future, to ensure you reap the long-term benefits.
You should invest as much time necessary to identify what it is you need from a new system and how the data migration will enhance your business. When polled, thirty-three percent of businesses identified poor system design as a challenge to migration; preparation, planning, expertise and a clear strategy will go a long way to prevent this.
Indeed, technical expertise is one of the best tools to ensure effective migration. Although you have an in-house IT team, it may be worth paying experts to manage your data migration, considering the potential disruption that an unsuccessful migration could cause.
Failing to test
Testing is an important part of your data migration process and should be done extensively, before your new system goes live. However, some data migration experts also recommend going live in stages.
Arvind Singh, co-founder and CEO of enterprise data solutions provider Utopia, believes that going live in one grand upload creates significant risk. Instead, he says: “You need a project timeline with multiple, iterative test loads along the way.”
It’s important to have a strong data strategy for your business and a data migration certainly shouldn’t be rushed. A good data management system can save you considerable resources and help you grow your business, while a bad one can damage anything from your reputation to your revenue.