dotData Hygiene Series: Part 1 – The Cost of Bad Data

As data privacy and usage slams into the agendas of lawmakers across the world, we thought it would be an opportune time to start our blog series on how organisations should be thinking about how they collect, use and maintain data. Tom Corbett, our lead on Best Practice, opens up with a introduction to the cost of ‘bad data’.

Notes Background 2

The data economy is exploding. Every single action and transaction we make within the digital space has a dollar/pound value attached to it. With that level of transaction and engagement comes inherent issues and risks with the associated customer data, all of which can seem like a minefield to most marketers.

So we figured this was an opportune time to start a ‘data hygiene and compliance’ blog series, as a resource for both best practice and future-proofing marketers’ concerns.

Our first in the series looks at the scope of the financial cost that poor data quality can cost your business.

A definition

While ‘data quality’ is a perception of an entire set of data, and its fitness to serve a purpose in a given context, bad data often refers to missing or incomplete entries. In the context of this blog, we can take bad data to refer to any set of data that has not undergone due diligence in its sourcing and maintenance. Typically this involves updating data sets, standardisation, and de-duping records to create a single view of the data – that is, the data warehouse.

Due Diligence

Before you invest in any new product, it is standard procedure to look at aspects such as cost, the quality of the product and how the supplier provides you post-purchase support.

However from experience, we have seen that the same due diligence isn’t being carried out with data lists, in particular how they are acquired and then maintained.

Bounce backs from an email marketing campaign are the first and most visible impact of bad data. However, what is the real cost?

The first is ‘Sender Reputation’ – rack up enough bounces and your sender IP reputation will get hit. Correcting this is a case of petitioning ISPs directly, or employing one of the myriad of whitelist consultancies out there.

Experian’s 2014 Global Research report, saw that over 54% of surveyed companies harvest data from call centres, and over 73% pull data in from their own website. However, only 38% of all respondents check data at the point of capture, yet 59%of all errors were identified as ‘human error’ at their root cause. The volume of data involved and the level of potential risk is staggering.

The Cost of Correction

Correcting data is a resource-intensive burden. The ‘1-10-100’ rule, quantifies the cost to fix a data error fairly well. At the time of entry that cost is approximately $1, factoring in average man-hours and competencies required. The cost to fix it an hour after it’s been entered is around $10. The cost to fix it several months later is over $100.

Ensuring that data is stored correctly against entries is another key consideration. The media is full of stories of personalised materials sent to deceased relatives and misspelled names. Every communication between brand and customer is an opportunity to nurture or destroy relationships.

With email marketing representing 56.8% of shopping cart traffic, and average B2C email marketing ROI smashing through 2500% (source: dma.org.uk), maintaining positive digital interactions with individual customers is more important than ever.

Experian’s Data Quality Infographic below, summarises a catalogue of risks associated with bad data, that their global survey identified.

The Data Machine infographic

Our next addition to the Data Hygiene series will look at a few different ways that brands can actively mitigate bad data.