A capability to make informed decisions based on data rather than hunches is very important, especially in unstable times. At Vincit, we typically help other businesses to transform their systems and processes to enhance their business. However, we are a business, too, and we have recently applied data-driven business transformation to Vincit. This post describes the key elements of why we did it, how we did it, and what we have learned so far.
Background
In July 2022 Vincit merged with Bilot and it was a major business event for both companies. I led the technology track of the integration process, as you may guess even though we had been preparing for the merger for months, not all details were in place. We did cover the basics: our employees were paid on time, and we did successfully invoice our customers. We made many changes to our internal systems and processes, and they generally worked. However, in many cases, the systems still were not linked to each other. Our system integration and process harmonization are still ongoing, and it may be good material for a separate blog. But we really needed a reliable overview of where our business was heading.
At the time of the merger, Vincit was still quite fragmented in terms of the systems: We just got a new PSA system, which was new for the majority of the users. We migrated customer data from Bilot’s two CRM solutions to Vincit’s master CRM. We had multiple accounting systems depending on legal entities and countries. We had not aligned most of the master data, including cost centers, or accounts in accounting systems. This lack of alignment caused a lot of manual work in the financial reporting, and it lacked in-depth business data, which we needed.
Making quick decisions
In 2022 we had official business checkpoints only when we published financial data to the investors. We had monthly reports, but especially in the changing global market situation we needed more frequent business checks and such that were not only limited to hard financial data. We wanted to get a set of business KPIs for the whole company. We wanted the data to be available initially 12 times per year, then 52 times per year, and finally, we aim at a daily decision-making process. Additionally, the data granularity was expected not to be at the legal entities that we had (8 legal entities at one point), but at the business areas that we ran the business in. This need resulted in a business transformation of the whole company and how we made business decisions. In order to provide the tools for our business to be able to make decisions in a more informed way and much more frequently, we established an internal project called Data Driven Vincit.
The Data Driven Vincit Project
The Data Driven Vincit (DDV) project started in early 2023. Initially, the project aimed at providing mainly OPEX forecasting. But in Autumn 2023 the vision was changed and expanded by our new CFO, Kimmo Kärkkäinen. We relaunched the project and rescoped it so that in addition to the usual financial reporting, it covered also sales, projects, consultants, and other targeted reports for our decision-makers to be able to see how their part of the business is doing. This was also the time when I was given the reigns of this project.
In many ways, it was a typical transformation project that we do for our clients, but the main difference was that we were transforming our own company. One could assume that, since we have great and capable consultants in-house, we used only our own people for the project. However, our great consultants are also the ones who are committed to customer projects, so we had a mixed team of internal and external experts. The team included specialists in:
- integration - to get the data from our various systems
- data warehouse – to transform and aggregate the data
- reporting – to visualize the data in reports
- platform – to take care of the deployment pipelines, security, etc.
- financial – to help with financial data calculations, e.g., cost calculations
- planning – to roll out and use a common planning tool for the whole company
We aimed at running in sprints, but admittedly, we ended up working on a flow of changing requirements in a more Kanban-oriented approach than actual sprints. The changing details of requirements were coming from the fact that the more business was getting visibility to the data, the more they were finding issues in our initial assumptions or asking for feature additions. So especially in Q4 2023 and Q1 2024 we were constantly releasing updates and improving based on the feedback. In Q2 & Q3 2024, the sprint planning and sprint stability has improved. Why I mention this is because even though we know exactly how a proper agile (Scrum or Kanban) project should look like, the reality may differ from expectations. The key objective of the project was to deliver relevant tools for our business, and we were making project adjustments based on that top priority.
From the decision-making point of view, the DDV project was sponsored by the CFO and our steering group included the whole Vincit Leadership Team. So, on one hand, we were able to get quick decisions and support, but at the same time, the expectations on delivering results were very high, too.
Technology
Our technological choices reflect the technology stack that we often work with for our customers. However, solutions can be relatively easily replaced with a different solution as well, dependent on the overall enterprise architecture of a company.
We decided to use Data Warehouse on Azure. We use Function Apps to fetch the majority of the data from underlying systems. The data is usually loaded at least once per day. The ETL operations are done with Azure Data Factory. And in the presentation layer, we use Power BI.
Data and processes
The main issues that we faced and had to solve during the DDV project were not technical, but were primarily related to the data quality and the processes governing the data. For example, many decisions that we made regarding our PSA system, which is Severa, were made during the integration project in H1 2022, and they aimed at merging data from three different systems of Vincit and Bilot. So, some of the choices made at the time had to be revised and improved. We managed to significantly reduce the hour types, which are essentially products that we sell as a consulting company.
Secondly, we had to clearly define and implement data masters for various data types. For example, our CRM which was a logical customer data master, was initially not integrated with our PSA. Therefore, we had to link the two systems and start working on data consistency, which included the removal of duplicates and the alignment of customer owners in the different systems.
Moreover, there were data dimensions, e.g. cost centers or profit centers, which were implemented across systems, but did not have a master system where they would be managed from.
Finally, as some of the systems did not have the technical capability to enforce certain process rules, we had to provide training and documentation for the users to follow what logically we agreed on. Also, in this area especially at first, the technical limitations caused users to perform manual work to ensure data consistency. E.g., our project leads had to link projects in two systems manually in a specific scenario, because of legacy integrations. Again, it is an example of a real-life challenge that we faced, and the temporary solution had to impact usability to get consistent data. Naturally, the permanent solution is a fully automated one.
Where we are today
When I write this post it has been about a year since the DDV project was re-launched. So where are we now? I dare to say we have provided a lot of tools for our business and basic needs have been fulfilled. The most notable areas are:
- In Finance, we have a full financial view of top and bottom lines not only on the legal entity levels, but at the internal business area and market levels. We can drill down to customer and project profitability.
- Our Finance Team introduced and ran a monthly forecasting cycle for all parts of the business. This was primarily a process change and our new forecasting tool merely eased the implementation of the change.
- Our Sales Team can see the Sales forecast and its realisation per business area, customer, or project
- Our Project Leads and business area leads can see project performance, project risks, and how they change over time
- Our Team Leads can plan and see the performance of their teams
- We managed to cover most of our CSRD requirements
This list is partially specific to our business needs as a consulting company, but many of the data sources could be replaced with a completely different business. We still have many items in our backlog and the more users use the reports, the more wishes or ideas they have.
We will continue improving the existing reports and work on our processes. I believe we have much better internal tools than in July 2022 and that they will have soon a visible impact on how our business performs.
