Big Data Business Intelligence Cloud Internet of Things Technology

Bigger is always better – let’s do big data!

Big data has been a “thing” for some years now and still many mid-sized and smaller businesses are figuring out if it is worth getting into it or not. The answer will most definitely be yes. To explain this rather radical statement let us define big data real quick.

In theory Gartner, Inc.’s definition “Big data is high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced insight and decision making.” gets right to the point. But I would like to use a little bit more practical approach. Every business uses data in some way, mostly there is a well-working process for it, which has been set-up once and runs for a long time without the need to tweak it a lot. Working with big data means to depart from this easy to handle and well serving process. “Big” starts when it gets difficult, and difficult is dependent on the company.

For a little bakery it could mean stepping up from giving you a free cup cake for every tenth purchase to offering you bread which fits to your personal health plan according to data from your fitness app and diet plan. For big players like Amazon it means to store and analyse enough data to predict trends and purchases before they even happen and to send out articles before you even order it.

Example for Data Usage in a Bakery
Example for Data Usage in a Bakery

There we go, the small shop is not able to manage your health data without new technology and external talent. So it is big data, while for other companies this piece of knowledge is not even close to what they store and work with in their CRM tools; not to speak of giants like Google or Amazon which manage more data every minute than all customers of the bakery produce in one year.

So the point of big data for going beyond, it is thinking of which data is out there and then comparing the volume, velocity and variety of the data and hence the expenditures which are necessary to analyse and use it to the advantage it would create in order to see if it is a beneficial investment. The thing is, in theory it will mostly be beneficial but during the implementation and execution of the new big data strategy companies tend to lose track and make mistakes.


Some of the mistakes are technical, e.g. using the wrong software and tools, some are preventable and some are not. Being aware of the following 7 rules will help you to execute your big data strategy successfully.

  1. Analysis: As just mentioned big data is always new and difficult (remember, this is a criteria for the “big”) and therefore needs a comprehensive and easy-to-adapt strategy based on a very detailed analysis. This analysis should be flexible and well-structured, so it is easy to refresh and expandable for the future. Include the numbers/data you already collect and think of how it is beneficial for you. Further you should analyse your own performance and the performance of your competitors to detect where your business is already strong and where it is behind the branch average.
  2. Strategy: Based on the analysis you are able to write your strategy. Start with your goals which may include to fill a performance gap or to strengthen a feature of your business or adding a new service. Whatever they are, the next step is to define the data which is potentially influential for each goal; think also about how you will analyse, update, present and use the data. It might be obligatory to mention but: plan your hardware, software and men power as well as internal and external human resources carefully and flexible as your data will most likely get more and more.
  3. Priorities: It is vitally important to also prioritize goals and the attached data to manage the big amount of knowledge you will get until all processes are in place. This step is part of your strategy but I like to mention it as separate bullet as you have to do this again and again. The data you collect will grow constantly and therefore you will have to prioritize data sets according to your goals repeatedly.
  4. Stability and Flexibility: within the KPIs you define you should have two groups. Stable once which are basically unchangeable, for example the number of new customers compared to the previous month. This number is a fact, directly tells you if your business is growing and you most likely won’t change this parameter much. On the other hand, to really uncover the full potential of big data, you also need what I would like to call “flexible KPIs”. These are KPIs which can change in two ways: the data which is flowing into them may change and/or the goal and performance they are attached to may change. An example for our bakery from the beginning would be data from your health app which the shop uses to determine if and which fitness products to offer. In the future the data may be accomplished by data from your shopping list app and may further be used to optimize the timely supply of ingredients for the bakery to offer you what you want the day you want.
  5. Patterns: Big data is nice but analysing every individual would be too much and would not help you to reach your goals. Therefore look for patterns across a group of individuals. Again decide if you want to you use the pattern static, to define a group based on it or if it is a flexible parameter which you evaluate and adapt periodically. It is also vitally important to recognize and compare patterns in the follower/customer base of your competitors. Returning to our bakery example this could mean that your competitors’ customers live less active and healthy life-styles. This might lead to the decision to offer some more cupcakes along your healthy stuff.
  6. Human Resources: In many cases you will need to hire new man power. The decision to make here is if you either hire an external agency or if you are willing to expand your team with a dedicated professional. There are pros and cons for both decisions. Long-term, under the prerequisite that you really use the data to generate ROI, a new employee might be the better choice, as this person will be able to understand your business needs better. On the other hand if your business is smaller or you predict that the amount a of data you will use will change quickly an external force will give you more flexibility. Best case scenario you are in the position to hire a person handling the implementation and strategy in-house supplemented by external experts and insights.
  7. Data Usage and Distribution: Last but not least after you have done everything mentioned above you will have to make a very important decision: Who will see the data and how will you deliver it? In a small company it might make sense to provide the full results of the big data analysis to all employees (e.g. in a weekly report). In a mid-sized or even bigger company you should take a different approach. Leaders should get access to a lot of important information to make the right decisions; this can happen in two ways: by hosting the information available for “pull” or by “pushing” the information to them. Again a combination is the best offering, letting them subscribe for information they want to get periodically while having access to all, up-to-date information in one central tool (e.g. intranet). The majority of employees, if not involved in the data analysis, should be able to access all information which influences their jobs, decisions and decisions of their leader on demand (pull) plus a monthly report with data that reflects success of the company and the several departments on a high-level.With this strategy you will accomplish two things:
    Your leaders get a big amount of helpful information to make smart and beneficial decisions.
    Your employees get all information they need to understand the decisions of their leaders, to understand goals and deadlines better and to compare the performance of their department to other departments.
    This will also allow and support every individual in your company to work more efficiently, to compare themselves to others and to have ideas on their own based on the data you provide.

Big Data is by definition never easy. If you will be successful depends on a lot of decisions and processes but the 7 tips here will help you get on the right track.

Do you have similar or different experiences, additional tips or do you disagree? Share your thoughts in the comments; I would love to hear from you!

For a short overview watch the great expainatory clip from the Harvard Business Review – The Explainer: Big Data and Analytics