So here’s my take on how the majority of companies (not all) see data quality (admittedly some do not see data quality at all, but that’s another blog post.)
Many see data quality as a percentage game – 80% good data is good enough – and there is plenty of merit to warrant this approach, but as I’ll try and explain taking data quality from maybe a 80/20 game to 95/5 game or 99/1 game or even 100/0 game is doable and the upside significant.
I can hear people screaming about 100/0 game but for some companies this is an achievable reality.
The most popular reason for not putting more effort into data quality is ‘we are growing happily, business is good and poor data isn’t hurting us.’
On one level this makes perfect sense, we’re always told to apply the 80/20 rule as way of managing our resources, and many conclude data quality is an expenditure project so why focus on this when we are growing, and have so many other important projects.
And that’s the assumption that is incorrect: that data quality is a cost efficiency project; data quality helps the top line by eliminating poor data that affects sales & marketing. If it was just a cost issue or a nice to have then the current corporate world wouldn’t be waking up to concepts like Data Science and Data Driven Marketing.
There’s a video on my blog (which is also included here) and on a LinkedIn post about the costs of poor data. It clearly identifies how poor data affects sales.
You’ll see there is a significant upside to fixing poor data in terms of lost opportunities and customer life-time value. You can see from graphs below, increasing data has a positive upside.
So, if you are a growing business focusing on data quality playing a 80/20 game, then you are missing sales opportunities in the remaining 20% of your data.
For most companies, the cost of fixing poor data is significantly less than the lost sales opportunities and substantially less than the Customer-Life-Time Values.
Whether the cost involves using a third-party supplier or having in-house data quality team or purchasing data quality/data management software, the costs will always be lower than the sales you are missing.
Watch the video again and create your own model of how poor data affects your sales. In most cases you’ll find a very desirable ROI.
Data Quality is a ‘Quick Win’:
1. Better data makes your marketing ROI higher
2. Your analytics more accurate
3. Your management reports more trustworthy
4. CRM users feel more confident
5. Invoicing more reliable
6. Costs wastage smaller
7. Protect your reputation
8. Increases sales!
There is another trend in the market place that reinforces the 80/20 game. Software providers and software-as-a-service providers for data quality try to automate the process of data quality. There are inherent problems in trying to reach 95% or 99% data quality with this approach. This leads to many organisations having data professionals to handle what can’t be automated.
Every company’s data is different and tends to have its own personality, hence making standard automated routines difficult to achieve high levels of success.
Where automation does work, is when the automation is bespoke to your data personality. This will cater for how data behaves in your organisation and you can truly get to 95%+ accuracy.
To many people bespoke automation may sound expensive, but it is still lower the lost sales numbers.
So the industry pushes, albeit unknowingly, the idea of 80/20 data quality when 95%+ is easily achievable when you connect data to sales lost.
When growth slows down you see many companies improve their marketing as they need to drive more sales, so they get more sophisticated.
One aspect of this sophistication is to improve data quality and making it a priority, so we can have better marketing.
Why wait?
Why wait until growth slows or poor data affects you in some other way that forces you to take action.
The cost is so much less than what most people imagine it to be. Everyone should make data quality a priority and create data that supports your business.
Whether your business dictates to you to improve your data quality (reactive) or you decide to proactively improve data quality, the time will come to improve at some point.
Better to anticipate this problem than wait for the problem to dictate to you.
In summary, the cost for 95%+ is disproportionately small compared to revenues hence there is no need to wait, waiting just loses sales.
