In 2012, the Harvard Business Review famously proclaimed that Data Scientist was the “sexiest job of the 21st century”. Okay, so they are probably tired of hearing it by now, but it seems those irresistible data darlings are still happy to embrace the title. In 2017, 88% of data scientists reported that they are “happy” or “very happy” in their job; six in ten agree that they have this century’s sexiest vocation.
Hold on a moment. Put aside the job satisfaction, the enviable salaries, the fantastic opportunities and being highly sought after by the great and powerful. Is manipulating data really a role to get people hot under the collar? How about after you consider all the dreary, menial tasks associated with cleaning and organising data in the first place? Because last time I checked, ‘data janitor’ wasn’t anywhere to be seen on the sexy job list.
In its most recent Data Scientist Report, data mining and crowdsourcing company Crowdflower found that 51% of data scientists do indeed spent most of their time collecting, labelling cleaning and organising data. By comparison, just 19% spend most of their time building and modelling data.
Typically, this time spent is inverse to the tasks data scientists say they like: while 60% say data drudgery like cleaning is their least enjoyable task, a whopping 78% say that building and modelling is a task they enjoy the most.
Therefore, for the most part data is dull. In the world of marketing (and likely many others), data preparation is more like living in a convent than on Love Island.
The trouble is, this tedious time-sucking stat-crunching and data cleansing is vital for just about everything modern marketers hope to achieve.
Without a trustworthy dataset for reliable insights, you won’t even get to first base, let alone go full on 50 Shades of Grey.
Think about all the sexy stuff that marketers want to do (if you can get your mind out the gutter for a moment). Say you want to use predictive modelling to segment your customers, build a chatbot to handle customer enquiries round the clock, or implement engaging personalisation options that speak to your customers on a one-to-one basis. If your data is garbage, none of this is possible. Well, it is possible, but don’t say I didn’t warn you when your email response rates drop through the floor, your campaigns fail, and your AI assistant starts calling your customers offensive names.
Fortunately, there are options. If you’ve yet to implement any sort of data quality regime, then a data quality audit will show you the status of your customer data, while a prescription of data cleansing and enhancement will get your house in order. These choices, while effective in the short term, aren’t ‘silver bullets’ – data never stays clean for long, hence why data scientists spend so much of their time doing boring jobs.
A more long-term solution(for data-driven marketers at least), would be through the implementation of a Single Customer View – a process that continually refines dirty data and merges disparate information to ensure you have an honest foundation for insights and analysis.
Whether your organization is lucky enough to have a sexy rockstar of a data scientist or not, the truth is that you need to be married to data quality before you can have a fling with personalisation or hyper-targeted segmentation. Get your data right and then you can get down to business.
Blue Sheep Data Quality Audits
Thousands of marketers use the Blue Sheep Data Quality Audit to assess and improve the health of their marketing data.
This free service can be used as a one-time assessment of your marketing data quality, or as the first step on a partnership with Blue Sheep towards an insights driven marketing team.