Data is worthless until it has had some level of analysis placed against it. For an organisation, data has historically been used to understand where it has been, where it is now and where it will be going in the future. This uses the time dimension of data- but there is a lot more that should be considered.
The place dimension, velocity dimension can add immense power to data analysis – particularly when combined with other contextual data. But these can wait for my other posts. Now what is this DATA INSIGHT all about?
A Short History of Data
The actual capability of any organisation or individual to operate successfully is based around having enough information available on which to make the right decisions. However, this information has to be based on data – and the data has to be accurate and complete enough to enable information to be gleaned from it.
There is a basic ‘Strategic Insight Pyramid’ that is used to show how data has to be acted on to increase its value. This shows how data needs to be dealt with to become information, from which analysis can create knowledge. Only from Knowledge can decisions be made – which then leads to strategic insights that can be used within the organisation.
That strategic insight is where the true value to the organisation lies – this is how intellectual property gets created and how the competitive actions, services and products that keen an organisation ahead of its competitors are progressed.
At a highly general level, organisations have moved over the past 30 years from having very little data being held electronically, through having proportion of data held in relational databases and a little of their less structured data held in enterprise content management systems, to a point where a vast majority of an organisation’s own data is being held electronically. What few organisations now use is data from outside their organisations. It is increasingly difficult to draw a line around an organisation and say “this is us”. Organisations are dependent on suppliers (and their suppliers), as well as customers (and in some cases, their customers). There is a wealth of data held along this value chain – but also even more, possibly useful, data outside of the constraints of the chain.
Just what is information?
So, data in itself is worthless. Information is what is uncovered once the data is understood. For example, knowing that the data is held held as a relational database, it is easy to understand the rows and columns the are hidden within the data. However, this has little value on its own – a row that is ‘Inedible Foods Ltd’ ‘250000’ ‘03032015’ ‘ABC23456’ ‘256’ hardly enlightening.Once the database structure is understood, then it may come to light that our customer, Inedible Foods Ltd, has a credit limit of $250,000 with our organisation, and that its last interaction with the organisation was on the third of March 2015, where it ordered 256 of stock number ABC23456.
Great – we now have some understanding of what is happening – but this is till not knowledge. All we have is one data point – we cannot see if Inedible Foods Ltd is going to buy more of ABC23456 next month; we have no idea where Inedible Foods Ltd is; we do not know whether it has a good credit history – we have no frame of reference to add any value to the information. This has been the problem with many data analysis systems – they can only analyse what is directly available, and tend to either give a historical snapshot view of what has happened or of what is happening now. There is a need to be able to access enough data sources to be able to provide better analysis of what is happening now, and what is likely to happen in the future.
Data has to be suitably aggregated, filtered and analysed so that it an be presented to the right person at the right time so that effective decisions can be made to create new strategic insights.
Gaining knowledge from information
It is increasingly unlikely that all required data will be held in one place. In the above example, the data is likely to have been pulled from an order system, which may not be linked to the CRM or other systems. By pulling multiple different databases together, more context can be pulled out.
We may find out that the organisation has been dealing with Inedible Foods Ltd for 20 years, that it orders every quarter and that it has never missed a payment. We may find a list of people within Inedible Foods Ltd that the organisation has dealt with. Context is now being applied to the data – and we are gaining greater knowledge by being able to apply time lines to what we know, allowing us to carry out a level of prediction-for example, we can be pretty sure that Inedible Foods Ltd will place an order within the next quarter- and based on buying patterns, what this might be. We are better positioned to know who to contact within Inedible Foods Ltd based on who ordered last time.
This enables a set of decisions to be made – what should be made ready to fulfil the order; when to contact Inedible Foods Ltd; who to contact within the organisation; send out an invoice and so on. We can start to put in place schedules for logistics. Value is now being added directly to the organisation – but it is still relatively low-level value.
Adding to the knowledge
Now let’s consider how other data dimensions can add to the value of the knowledge already obtained. Let’s assume that the address for Inedible Foods Ltd within the master record is ‘1 Inedible Foods Way, Sometown, Anywhere, AA1 3AA’. We now have a geographic fix for the office. However, we know that Inedible Foods, Inc is retailer, and that it has multiple shops around the country. Is this valuable to us? It could be – it could well be that Inedible Foods Ltd has multiple distribution centers. Is there value in delivering directly to each distribution center, rather than to Inedible Foods Ltd’s main central distribution centre? Will Inedible Foods Ltd pay extra for that? How about direct delivery to each outlet – will that enable better ‘just in time’ (JIT) inventory management for both our organisation and Inedible Foods Ltd? That could have great value for both organisations – and just requires a little knowledge of where everything is and suitable mapping software to compete optimal vehicle routing.
We do, however, need to be able to easily visualise this. It is not that easy for an individual to look at a list of addresses and figure out the best way to arrange logistics to visit the addresses in turn. Layering the data directly onto a map makes this much easier for the individual to understand. The data already has geographic context within it – each address has its own postcode. It therefore becomes easy for the right analytics systems to take that data and layer it over a map – and then doc carry out further analysis and advise on the best route for trucks to take to minimise the time and fuel involved in getting to each destination.
This can then be combined with other datasets, such as inventory, to ensure that outbound loads are maximised. By analysing customer needs, it also makes it possible to offer them reverse logistics – do they need something transporting from their site to a point along your vehicle’s return trip? If so, a charge can be made for this that offsets the cost of outbound logistics, while adding extra value to the customer relationship. With inventory, being able to add a dimension of place to each item can be valuable – for example, we may not have what is required to fulfil an order in the warehouse we have direct responsibility for, but is the item in any other warehouse worldwide, and how long and at what cost would it take to get the item either to our warehouse or directly to the customer? With ‘big data’ coming through, do we have access to live inventory information, such as where a specific pallet or item is in its journey? If so, can we make this available to our customers so that they can self-track items and so cut down on enquiries to our help desk?
Now the true value of data is being made visible. We have passed beyond just being there as information; we have added value to it so that it becomes knowledge that business nations can be taken against. The data has reached a position where it is not only helping our organisation take informed decisions, but it is also giving us capability to help our customers and suppliers – making them more likely to maintain high value relationships with us.
The application of context
So far, we have considered the direct analysis of available data on a single entity – in this case, Inedible Foods Ltd. However, to become a truly intelligent organisation based on analysis of all available information, we need to extend our reach and add further context. For example, Inedible Foods ltd is player within a specific retail market. By brining in information about others in the same market, we may be able to both create additional value for ourselves and for Inedible Foods Ltd.
Maybe we can see that two of Inedible Foods Ltd’s competitors are merging, and that this will result in a major overlap in the new entity’s stores. By sitting down with Inedible Foods Ltd, maybe we can help them to understand where to put in more investment where the new entity’s outlets may be closed down and to rethink plans where existing outlets are likely to be increased in size. Maybe we can work with Inedible Foods to come up with an optimised delivery system where we deliver to major shops and they take some of that to smaller shops themselves, and w charge them less for the delivery.
We may also be tracking the market and find that there are increasing problems in the market that Inedible Foods Ltd operates in. Maybe we should be looking at tightening our credit limits on them – or maybe not, if the information only applies to geographical areas where Inedible Foods Ltd does not operate. It is pretty obvious that there are three main aspects to ensuring that an organisation can gain optimal value from its data assets – having access to as much real and correct data as possible; being able to understand the context of data with regards to place. This combination enables fast analysis of data to provide the information to a visualisation engine in the correct manner so that knowledge can be made visible to a person, enabling timely strategic decisions to be made – so completing the Strategic Insight Pyramid.