Big changes are coming in the New Year with the launch of our new brand!   - If you have any questions in the mean time, drop us a line!

+44 (0)330 313 1000

by Graham Cookson
in Big Data , Hadoop
on 24 September 2015

10 tips to harnessing your (internal) big data: part 1

When it comes to looking at the data within your company, there are no right or wrong methods. What works for one company, may not work for another. But there are best practices and guidelines that every company should consider when looking to get the most out of their internal big data.

At Digital Contact we have put together a list of 10 tips for any company in any industry that we consider to be of high importance when harnessing and getting the most value out of your internal big data.

Note: By ‘internal big data’, we are looking at the information naturally generated by the day-to-day operation any organisation, such as (but not limited to): sales, purchase orders and transactions.

1 - Richer datasets mean richer data 

It sounds simple and, really, it is simple – richer datasets mean there are better chances of finding useful and important insights. Often times, when you have a rich dataset, the insights will jump out at you.

The way to ensure you build richer datasets are threefold:

1. Collect more data

Don’t discard or ignore the data that comes your way. Sometimes the data you receive might seem pointless at the time, but in the future it could become valuable to your business.

It’s a lot simpler to collect that data now, than ignore it and find that you do need it and have to try and collect it all over again.

2. Create more data

Not every company is the same and there are those who obviously create data on a daily basis (tech companies for example). Then there are those businesses that you might assume don’t really handle data and, therefore, don’t create data or need to create data. You’d be wrong.

It doesn’t matter what your business is, data can be economically created/generated that can be incredibly useful.

For example, imagine you own a small shop (a bakery for example), which you might assume doesn’t need to use or create data in the same way as a tech company. But you can still create and use data to benefit the business - such as asking customers how they found your bakery.

By asking this simple question, you are instigating data generation and the results can be used to strategically position your marketing efforts in the future.

3. And store it!

This is a part that’s so painfully obvious for any big data company, yet not so obvious to everybody else.

Quite often companies will collect and create data and then simply discard it, even after finding out insights – with the assumption that because the value has been found already in the data, they don’t need it anymore.

As we said at the start: richer datasets mean richer data. Keep that data, the value will maintain for a long time and often grow as your datasets do. 

2 – Examine your data closely

While insights can seem to jump out at you, don’t be disappointed if they immediately don’t. Sometimes they are hidden deep within the datasets, such as: sales receipts, third-party transaction logs, website analytics, marketing surveys, etc.

Be sure to spend time investigating your data carefully. It’s not a waste of time.

Understanding your data will allow you to grasp its potential and it’s possible you’ll find nuggets of gold hidden within that can help your business in ways you didn’t realise.

With persistence, you can find valuable insights into both opportunities and threats you might otherwise have missed.

It pays to find meaning from meaningless data. 

3 – Experiment!

Everyone’s data creation and collection is different, as is the value you can gain from it.

Sure, similar companies with similar goals are a good guideline and it’s good to take suggestions and examples from competitors and collaborators, but the best way to find real value is to experiment and experiment often.

Don’t experiment once and walk away with those results thinking you won’t find anything else – continue to experiment and play around with your data.

By doing so, you will learn how best to use your data, what works best for you and, ultimately, allow your business to evolve. 

4 - Go big or go home

Use big datasets as often as possible. There’s a reason it’s called ‘big data’ and there’s no reason to not utilise large datasets.

Technology has advanced rapidly in the last 10 years or so. Not so long ago, the thought of using large datasets would have sent a shiver down the spine of any statistician and many companies would have relied on small datasets to base their judgements on.

Because this old practice, companies will often take samples of datasets from within their company and get very little out of it.

The evolution of big data technology means that you can now analyse large datasets and see a broader picture with relative ease.

5 – Look at multiple data sources

Note: This step could be considered to be more advanced and is perhaps not something that a company new to big data should just jump into. Although, when you do finally take this step, the benefits can be far reaching.

Using one source of data is good and can lead to valuable insights. However, by utilising as many sources as possible and cross referencing them at the same time, it’s possible to find the true value in your datasets.

You need to ensure that you have a big data analytics/business intelligence tool that can blend data across multiple data sources.

It’s more advanced, especially for smaller businesses, but if you find yourself faced with multiple sources, it is often the best practice (these days) to broaden your data view to deliver richer insights.


Part two of our 10 tips to harnessing your (internal) big data >>>


Disclaimer: This article is for informative and guidance purposes only. It does not offer direct advice on managing big data.

For more information and guidance on how to harness your data, please contact Digital Contact's Big Data Consultancy team.

Email:, or call: 0330 313 1000.

Liked this? You may also be interested in: