Small Data: harness your micro-data to manage your SME
Small Data, Big Data... while we hear a lot about the latter, the former is making its way more discreetly.
Yet the two concepts are not far apart.
Small Data even offers better prospects for the management of very small businesses.
According to MyDataModels, a pioneering French start-up, small data accounts for 85% of all data collected.
Let's take a look at this promising trend.
What is Small Data?
Small Data: definition and objectives
Small Data represents all the micro-data, or micro-information, collected daily within a company, via :
- files :
- Excel spreadsheets,
- team schedules, deliveries, projects,
- internal studies and reports,
- minutes,
- photo and video files, etc. ;
- applications:
- diaries,
- electronic mail,
- instant messages,
- social networks ;
- operational software such as CRM (customer databases);
- physical or digital sensors.
Real decision-making levers based on objective, quantifiable and measurable criteria, this data accessible can be used to study and then optimize the productivity and efficiency of all departments, in particular :
- human resources,
- sales
- marketing
- logistics, etc.
Small Data vs Big Data
What's the difference between Big Data and Small Data?
With the explosion of the Internet, users have never stopped creating digital data (videos, photos, texts, etc.), sharing it via different channels and storing it in the cloud.
It is this explosion in volume, but also in the variety of content and the need for speed, that has led researchers to find a new way of storing and analyzing this data on a digital basis: the phenomenon of Big Data.
It applies above all to the financial and sales sectors, but also to telecommunications, health, industry, government and so on.
With the development of Artificial Intelligence (AI), large corporations, notably the GAFAMs, and their data scientists can exploit this massive public and private data to guide their strategies.
Small Data, on the other hand, runs counter to this trend towards folie de grandeur: sometimes, the little data in the possession of the smallest structures can be quite sufficient to improve their performance.
Better still, if properly selected and reduced to a minimum, the resulting predictive analyses can sometimes be more accurate than the most complex algorithms.
That's why it's aimed at business experts who are confronted with a wide variety of data, but don't know how to exploit it on a day-to-day basis. We're thinking, for example, of :
- product managers
- marketing managers,
- financial managers.
In the following video, HEC professor Olivier Sibony sums up and illustrates this phenomenon very well:
What about Smart Data?
It's not about the data itself, but about the way it's used: by focusing on interesting data, decision-makers no longer drown in a mass of useless information, and use only the data that proves to be relevant and usable in the context of a specific problem.
In a way, Smart Data is the result of an initial sorting of data:
- from Big Data
- generated by the company itself, when the data is substantial.
Increasingly, this task falls to the Chief Data Officer.
Small Data: application examples
Small Data can be used by human resources departments to manage talent or improve quality of life at work (QWL).
For example, ManPower plans to analyze the frequency of use of internal communication tools to detect influential employees or possible drops in motivation, requiring the deployment of a retention strategy.
Another case in point: as part of customer journey optimization, sales and marketing managers can use the Small Data at their disposal, in particular transactional data such as average basket and place of purchase - physical or web - to :
- understand blocking points,
- know where to focus their efforts.
How can you make the most of your Small Data?
Analytical tools
The prerequisites for making good use of your data are :
- standardize data collection,
- centralize them in a single tool.
Data management and analysis platforms include :
- Data Management Platforms or DMPs, data analysis solutions for collecting, reconciling and unifying customer data, for personalized marketing actions;
- dashboards generated by :
- ERP, marketing, CRM and HRIS software,
- Business Intelligence (BI) tools, which connect to all these different data sources,
to collect and track the evolution of several performance indicators (KPIs), selected in advance according to your objectives (hello smart data!).
But beyond data analysis, these tools can't make predictions from data. Hence the emergence of the following solutions.
Predictive tools
Machine learning can be very useful here.
This concept is based on mathematical and statistical approaches that enable computers to learn automatically from collected data, and thus solve tasks autonomously.
With this in mind, MyDataModels has unlocked the technology of Big Data's evolving algorithms to make them usable on a smaller scale.
Even from small datasets, Artificial Intelligence can :
- extract value from data, by transforming your expertise into relevant and interpretable data, simply and directly by business experts;
- produce automatic predictive models to help companies improve their processes.
Result: the TADA by MyDataModels solution makes Small Data "talk" to researchers, medical experts and manufacturers, helping them to make predictions based on automatic learning models.
Small Data + Big Data?
If you have the resources, it makes sense to exploit both.
Big Data provides general trends in your industry, consumer habits or the behavior of your typical customer, while Small Data puts them into perspective with your expertise, your business data and your own experience of the business.
In short, with Small Data, data is exploited through technology, without neglecting human vision.
Article translated from French