Overview
Vendasta is a Canadian-based technology company. It helps small and medium businesses to manage different digital channels. Vendasta provides a platform where local experts (digital experts or IT companies) can get marketing and sales tools, outsourcing services, task tracking, invoicing, and business training.
Problem
Vendasta created an automation system to minimize the manual work of both its customers and its internal teams. Automation systems are powerful but can quickly overwhelm a company's infrastructure if something goes wrong. Vendasta needed to understand the following before releasing the feature:
- How many users will there be?
- How long will it take those users to adopt automations?
- Are those users internal Vendasta teams, Vendasta's customers, or both?
- Is there a risk that the automations will rapidly activate paid features or overwhelm Vendasta's infrastructure?
- Are there benefits or costs that Vendasta missed and should explore further?
Vendasta needed a way to quickly, yet carefully answer these questions. The company also didn’t know how the platform’s user profile would look like. Vendasta had only predictions but needed qualitative research.
Solution
Vendasta developers built a system dynamics model with AnyLogic software to explore different scenarios. The modelers designed several diagrams in the model including partner adoption and retention, feature uptake, automations running, etc.
System dynamics modeling gives the opportunity to observe a lot of dependencies and make the right findings for a real system application. Simulation with AnyLogic is good for what-if analysis and risk management even without historical data.
The AnyLogic software supports the design and simulation of feedback structures, such as stock and flow diagrams, and array variables (subscripts). Frequent patterns may be saved as library objects and reused within a simulation model or across different models.
The company analyzed how many partners would use the platform and how many automations these partners would create and run. Vendasta divided partners into "good-fit" (9-15%) and "poor-fit" (1-15%) and explored the adoption scenarios.
Vendasta wanted to include in the models all stakeholders: development, support, sales, and R&D teams. The modelers looked at processes in the sales and support teams to determine how much work could be avoided by using automation. The R&D teams used automations to implement features, so it was important to know what infrastructure and time they would need to accomplish it.
Results
The simulation modeling outputs related to external automations were almost the same as the initial predictions. For example, the quantity of running projected automations was 350, and the simulated value was 361. The number of projected automation users was 400, while the actual amount was 319. The difference in both cases was only around 3%.
Vendasta also considered the benefit of internal automations. The company managers realized that even if the internal automations didn’t save much time or money, they could affect employee satisfaction and productivity. For example, the sales team had daily routine work which automations could remove.
Also, using the simulation results, Vendasta analyzed the work of the support team and prepared specific detailed documentation for their needs. This might improve the quality of employees’ lives and their performance in the company.
Watch the video about the case study presented by Vendasta at the AnyLogic Conference 2021:
