Before talking about BI project, a quick reminder of definition : The Business intelligence (BI), which is also called business intelligence And sometimes decision support system (DSS), consists in placing a set of technological resources, methods and computer tools at the service of decision-making within a company.
For “informed” management, i.e. an approach that really allows the company to be able to rely on reliable data and relevant analyses to make tactical, strategic and operational decisions, well Supervise your BI project is fundamental.
The essential functions of the BI project decision chain
Before exploring best practices for a successful project, it is important to remember that a business intelligence project must respect the steps that make it possible to structure the decision-making chain:
- A collection/feeding phase: operational data (internal to the company) must be collected in a relevant manner through a ETL process (Extract Transport Load), i.e. retrieved from the various sources available, filtered and adapted for use for decision-making purposes;
- A storage and modeling phase: the data is structured, centralized and made available in the Datawarehouse; the latter must be non-volatile, business-oriented, historical and integrated.
- A retrieval/distribution phase: it is necessary to be able to return the data and to offer easy access to it by taking into account each profile and business need, whence the existence of Datamarts. This stage includes reports, statistics generated, ad hoc or mass reporting tools, dashboards, tools for navigating OLAP cubes (or hypercubes)...
- An analysis/exploitation phase: from the reports generated and data, the end user must be able to analyze the information and draw conclusions from it. This phase may include datamining to explore possible correlations, multidimensional analyses, performance analyses...
What does it mean to “succeed in a BI project”?
Concretely, a successful BI project will make it possible to increase the company's or customer's revenues, to get ahead of the competition at various levels, to optimize internal businesses and processes, but also to facilitate decision-making and the choices of managerial throughout the year.
4 important steps to successfully complete your decision-making project
1) Define your needs, objectives and the scope of the project IB
The success of the BI project will in particular rely on a clearly defined business strategy. This means having a clear vision of the future and being able to work on the basis of goals. Realists, clear in the short, medium and long terms, not to mention the phase ofStudy of the existing.
This strategic control will make it possible to anticipate and to manage deadlines. Business intelligence must adapt to the pace and projections of activity, not the other way around.
It is also important at this stage to properly identify What types of information decision-makers, project managers and other key roles will need to improve their data management.
Of metrics precise must be determined keeping in mind the strategic framework of the project: the relevancy of these KPIs is fundamental to avoid going blind based on “global” or industry-specific performance indicators.
2) Choose your methodology, with the possibility of opting for an agile approach
Agile or V-cycle? The choice of a methodology adapted to society is essential for the success of the project. In particular, it is necessary to take into account the speed of change in needs and priorities.
Too often, BI projects are developed “apart” from the life of society, without engaging all the strata involved at the various stages of reflection and implementation, which means that a lag that is as time-consuming as it is counterproductive occurs. Opt for a agile approach will then make it possible to gain in reactivity, to opt for frequent deliveries and to limit iterations, where a traditional approach will involve stopping all the details of the project from the start.
For other economic actors, the BI project will be based on a more traditional approach such as that of V-cycle.
Regardless of the methodology adopted, a process of governance of data quality must absolutely be put in place, to ensure in particular that the data is reliable, well sorted and up to date. This dimension is all the more important as differences between data can be frequent.
3) Choose your business intelligence tools and share your action plan
Successfully carrying out your decision-making project requires BI tools adapted, but also exchanges and a consistent training of work teams who will be involved, for optimal adoption within the company: users of the tools in question, details of shares coming soon, identification of resources And infrastructures needed.
Again, the project manager will play an important role. In particular, it will contribute to guide the choice of BI solutions based on the company's strategy and specific expectations... and not the other way around!
The action plan should also specify the hierarchy and tasks of each person in terms of monitoring and implementing the BI project.
4) Organize reporting and (well) transmit information
So that the people in charge of decisions can really decide, accurately, have access to information that is clearly organized and presented. It is not an aesthetic whim: clear dashboards, diagrams, readable figures and other reporting formats will make it possible to find your way around and prevent the BI project from resulting in a cluster of data that is rarely or poorly consulted.
Cela implique plusieurs choses : une perspective à garder en tête lors du choix de l'outil, une formation technique adéquate des personnes amenées à l'utiliser et le consulter, ainsi qu'un design clair qui donne du sens au contenu des rapports.
Par exemple, construire un tableau de bord fonctionnel est une question d'équilibre. Du plus simple au plus complexe (qui va par exemple agréger les données issues d'une multitude de sources, CRM, Google Analytics, ERP, etc.), il doit être suffisamment fourni pour donner accès aux chiffres et informations nécessaires à la prise de décision, sans tomber dans le trop plein de complexité. Pour décider de manière éclairée et passer à l'action, ce type de rapport peut aussi inclure des commentaires, des recommandations pour l'action et une estimation de l'impact de cette dernière.
Au final, cette étape de l'accès à des données pertinentes et organisées est donc crucial, faute de quoi le projet BI ne pourrai déboucher que sur des décisions peu ou mal informées, sur l'absence d'action ou sur des mésinterprétations qui peuvent directement menacer les objectifs recherchés, et avec eux les performances.
Chez Alphalyr, nous mettons la donnée au coeur des problématiques clients. Nous ajoutons aux KPIs classique des données de contexte comme la météo, les opérations commerciales, le carnet numérique, la concurrence, les objectifs. Le contexte est indispensable pour rendre un reporting de ventes actionnable. Chaque matin, chaque manager reçoit un reporting actionnable et personnalisé. Nos reporting off et online permettent à chaque manager de se consacrer aux ventes chaque jour.
Pour en savoir plus, demandez une démo !
