Toduba started from a conversation about a specific, mundane problem: every October, HR leads at companies with 150 to 400 employees sit across from a benefits vendor and renew roughly the same plan they have had for the past two or three years. The conversation focuses on pricing and maybe one or two category adjustments. The underlying allocation, what portion of the budget goes to commuter subsidies versus wellness versus learning, stays almost exactly the same.
This pattern is so common that it reads as normal. We do not think it is normal. We think it is a problem that produces real cost, measured not in euros spent above budget but in euros spent on the wrong things year after year.
The practical reasons it happens
The first reason is time. The annual benefits review window at most mid-market Italian and EU employers runs four to six weeks in the autumn. During that window, the HR team is also managing performance cycles, budget planning requests, and often a parallel Q4 recruiting push. The mental bandwidth available for a genuine reassessment of the benefits mix is limited. Renewing what worked well enough last year is rational under those constraints.
The second reason is data access. Claim history is typically held in the benefits platform or with the third-party administrator. Pulling it in a format that allows category-level utilization analysis by workforce segment requires either a technical extract or a specific reporting request to the vendor, neither of which is fast or simple. Most HR teams have aggregate utilization numbers, if they have anything at all. Segment-level patterns, the data that would actually support a redesign decision, are not typically surfaced in standard reporting.
The third reason is risk aversion. Changing a benefit category means some employees may lose access to something they value, even if most employees were not using it. The complaint from a vocal minority about a category change is more salient than the diffuse benefit to the majority. HR teams, rationally, weight the visible cost of a change against the invisible benefit of a better allocation.
These are not failures of professionalism or competence. They are structural features of how benefits management works in most organizations below a certain scale. At large enterprises, there are specialist total-rewards teams with dedicated analytics support. At companies with under 400 employees, one or two HR generalists are managing the same complexity with a fraction of the resources.
What the renewal instinct costs
The annual renewal instinct costs money in a specific way. Benefits categories that were well-calibrated when they were introduced drift out of alignment as the workforce changes. A commuter subsidy that was appropriate for a workforce commuting from Turin's outer districts in 2021 may be less appropriate for a workforce that hired heavily in 2023 and 2024, where a larger proportion of employees live centrally or work flexibly. The category persists not because it is still serving the same population well, but because no one has had the time or data to notice the drift.
The cumulative effect over three or four renewal cycles is a benefits mix that is a historical artifact of past workforce compositions rather than a current reflection of what employees need. The budget being spent is real. The value being returned per euro is lower than it should be.
The problem Toduba is working on
When we started building Toduba in early 2025, we focused on the data-access problem as the primary constraint. If an HR team cannot quickly pull two years of category-level claim history broken down by meaningful workforce segments, they cannot have the conversation about whether the current allocation is right. The renewal instinct is partly rationalized by the fact that the data to challenge it does not exist in a usable form.
What we are building is an analysis layer that takes claim data from the platforms where it already lives and surfaces the patterns in that data in a format that fits inside a normal annual review workflow. Not a two-week analytics project. A view that an HR lead can work with in the same afternoon they are preparing for the vendor conversation.
We are not building a system that replaces HR judgment. The model surfaces where the data suggests the current allocation is underperforming and what reallocation options would improve expected utilization within the same budget. The HR professional looks at those options, applies their contextual knowledge of the workforce, and makes a decision. The model narrows the option space. The decision is still a human one.
What this does not solve
We want to be precise about what we are not claiming. A better analysis layer does not solve the time problem. If the HR team has four weeks and seventeen competing priorities, the time available for a benefits redesign conversation is still limited. What the analysis layer does is make the time that is available more productive. Instead of spending two days pulling and cleaning data before any analysis can start, the HR team can spend two hours on the analysis and two days on the decision.
It also does not solve the organizational dynamics around category changes. An HR lead who wants to move budget from a category that 8% of employees use to one that would serve a larger segment still has to manage the conversation with the employees who will lose access. That conversation is a real part of the work, and no amount of better data makes it go away.
What we think better data does is make that conversation clearer. When the decision to change a category is supported by two years of utilization evidence across workforce segments, it is a different conversation than one that is based on a hunch or a survey response. The evidence does not remove the friction. It gives the HR professional a defensible position from which to have the conversation that needs to happen.
That is the problem we started from. We think it is solvable. We are early in the process of solving it.