How to unlock commercial value from your data.
Start asking the right questions of your data and progress to more sophisticated analytical techniques…
In my fourth post in this six-part series exploring ways to extract commercial value from your data, I’ll explain how to do just that – unlock the hidden value of your data.
I’ve already shared what I believe to be the fundamental traits of a successful data-driven organisation. As I explained in previous posts, a strong data culture and strategy, along with robust data auditing and management processes, are all essential.
However, businesses that implement the right culture and processes might still fail to extract their data’s full potential. This comes back to one of the key principles of analysis – creating meaningful insight isn’t about the data you hold, but the questions you ask.
The better your questions, the more insightful your analysis will be. And in order to form good questions, businesses need to democratise their data. They must simplify it and remove any barriers, so anyone can access the data at any time and ask relevant questions.
Below, I’ll outline how you can make data available across your entire business, leverage department knowledge and expertise to ask the right questions – vital in moving from basic automated reporting to generating insight and then, finally, more advanced techniques, like predictive analytics.
Democratising data.
In my experience, the way data is used within many businesses leads to inequality. Accessing data requires a level of knowledge, experience and even something as simple as the right level of system access. An organisation must trust that an individual is adhering to data protection policies and regulations too.
People who have access to data are at an advantage. Senior management teams are hungry for analysis and insight to make decisions – so much so that they’ll often request information directly from the analytics team without the relevant marketing, sales or service department's involvement or even knowledge.
For me, this isn't right. It misses something crucial I’ve highlighted multiple times in this series of blogs – context. If analysis and insight bypass the individuals and teams closest to the customer, a valuable step in the process is being overlooked – the contextual understanding of the impact of data-driven decision making.
Collaboration between departments is key, which is why businesses need to prioritise democratising their data. But how?
Make the inaccessible accessible.
Many of us will have experienced the inaccessibility of data – whether it's caused by legacy systems which don’t talk to each other, restricted access rights for certain individuals or raw unrefined data that’s impossible to interrogate.
Removing these obstacles and making data accessible across an organisation is an important first step. Data warehousing platforms such as Snowflake provide an environment for organisations to send and store data in a structured and accessible format.
Simplify complexity.
Many of the methods, models and systems used to store data are often illogical for a non-technical individual. A huge amount of effort from analytical teams goes into structuring, preparing and validating data to make it fit for further analysis.
On average, analytics teams spend up to 45% of their time getting data prepared before being able to analyse it. Without ‘data wrangling’ – as it’s called by analysts – data is unstructured and won’t correlate across multiple sources.
Set up interactive reports, dashboards and visualisations.
Along with unearthing insight of their own, a key part of an analyst's job should be to enable insight creation across an organisation.
By turning raw data into interactive reports, dashboards and visualisations, other departments can interact with and query the data, despite having little technical expertise. This opens up the data and the possibility of anyone in the business generating meaningful insight from it.
Asking the right questions.
As I touched on earlier, creating meaningful insight isn't about the data you hold, but the questions you ask.
As someone who’s worked as an analyst, I feel I can say that the best questions don’t always come from analysts. Marketing and sales departments are closer to the customer and hear positive (and negative) feedback about the brand. It’s why, as I wrote in my first article, if the only people who understand your company’s data – and its value to the business – are analysts, you’re failing.
To start asking the right questions, you need to have made your data accessible, simple and interactive. At this point, you don’t want to open it up to thousands of questions and have people going off on tangents. Although you have democratised your data, you still need to retain some oversight.
Helping employees come up with the right questions is a good starting point. At Levo, we developed the framework below, which should provide you with some food for thought.
The success of data democratisation and asking questions still comes back to what I have spoken about over the past few weeks - collaboration. If your marketing and sales teams are drawing their own conclusions from the data and not working with the analytics department, you still have work to do.
Predictive analytics.
As your organisation becomes more data-driven, you will naturally start moving towards more advanced analytical techniques.
The vast majority of the time, data is used to analyse the past – what happened, why it happened, when it happened and so on. As useful and insightful as this can be, data becomes much more interesting and valuable when it's used to predict the future instead.
One application of advanced analytics is predictive analytics – in other words, making predictions of future outcomes based on historical data. It relies on statistical modelling and machine learning techniques to create these predictions.
Companies like Amazon, Google, and Facebook use a predictive analytics technique known as propensity modelling to drive recommendations and their advertising platform. This takes a prospect or customer’s profile information and historical interactions to predict their next move.
At Levo, we have applied the same methodology in the B2B space. We leverage data to help brands identify their best sales opportunities with over 95% accuracy. You can read more about this here.
Implementing propensity modelling does require a high level of technical expertise, but the involvement of marketing and sales teams is still essential. As well as using historical data, an effective propensity model should consider; macro changes in the market, qualitative insight from the sales team and your organisation’s overarching marketing and sales strategy.
These inputs will allow you to model different scenarios and test how market forces and changes in strategy may affect your customers’ propensity to buy.
Final thought.
To reach the point where you have fully democratised your data – and have your organisation asking meaningful questions – takes time and investment. But without the right culture, strategy and processes in place, this investment will be wasted.
With more businesses becoming data-driven, implementing new predictive analytical models and using data in smarter ways, organisations need to ask themselves – are they doing enough not just to keep up, but to stay ahead?
Next week, in a video I’ll review the advice given throughout this series of blogs and share examples of businesses that have demonstrated a number of the points I’ve made.
Want to hear more?
To receive regular insight, updates and advice, simply sign up to our newsletter.