The Process
The tendering process is familiar for most companies, either for buying or selling goods and services, leasing, or as it was with Skanska’s case, for selecting property sales agents. Sometimes the process can get pretty complex, and that is why Skanska, the leader in construction and project development, decided to optimize it and created an online application system for agents’ tendering from very scratch. To complete this goal, Skanska enlisted the help of Uptime Development.
But let’s start it from the beginning. Skanska faced a problem in the process of selecting divestment agents: the process was not efficient, difficult to organize with many stakeholders on Skanska’s side, and the data received from the agents was difficult or even on occasion impossible to compare. As such, it became clear that a tool was needed to digitalize and standardize the process, to eliminate the number of received offers, and to save time.
Throughout several rounds of consultations and brainstorming sessions, Skanska decided that the best way to improve its processes was to develop an online application system with the help of Uptime Development. The app consists of two environments: the admin-side for Skanska’s employees, which allows to create and edit new tenders, and the public-facing side for the divestment agents to fill out a form and submit the tender.
Better user experience
“Once all forms are completed by the agents, Skanska can use their admin environment to review and rate the offers. Thanks to the fact that all of the agents are filling out the same form, it has now become much easier to compare the offers and the scope of information.” explained Rafał Szybowski, Uptime Development’s team leader.
This, however, is not the final form of the system. To remove all subjectivity from the process, the long-term goal is to use AI and machine learning to suggest the company to work with. To achieve that, the application needs to be trained to understand all the factors that contribute to Skanska selecting the winner.
“So far F2F meetings and receiving complex offers, which have to fit certain selection criteria, have been the base for selecting the company to work with. However, to make proper and semi-automated decisions, the system must understand which aspects were key for picking the agent,” Szybowski added.
Currently, the system is registering everything that is being done, to construct the necessary database for building the model. “Once that’s completed, we can create a system that will generate automated reports and suggestions. Skanska’s employees can review these reports to make sure that the AI picked the most suitable offer and once they confirm that, the contract can be awarded,” explained the team leader.
“The project is innovative and has not been implemented on the real estate market so far. Some agents adapted quickly to this new form of working and some faced challenges and were questioning this new approach. We believe though that the collected data will create a knowledge base for us that will be used in the next tender processes and in the daily work of our team. Thanks to the standardization of the data format, we will exchange the same scope of information with all agents and thanks to that the process will be completely transparent,” underlined Adrian Karczewicz, Head of Divestments at Skanska.
“The work on the project gave rise to several interesting and complex challenges, mostly related to updating Skanska’s business processes and AI applications,” said Szybowski. “But thanks to the very high commitment from the client and solid analytical work, we were able to develop a system that serves Skanska’s needs today, but also offers room for further improvement.”
