From Fragmented to Fully Synced: How I Approach Vendor Partnerships as a Product Leader

If you didn't catch our recent webinar with VETRO, we spent the time talking through a problem most ISPs know all too well. Connections between systems that technically exist, but do not work together in a way that actually helps teams move faster or serve customers better. The session focused on moving from fragmented systems to connected ones, using the relationship between an OSS or BSS and a fiber management platform as a practical example. What follows is a short recap of that conversation, along with some additional perspective shaped by my time running an ISP and now building software for them. 

Before joining Sonar, I was head of operations at an ISP. Like most operators, we did not start with a sophisticated network design, or a well-calculated vendor stack. We layered on solutions as the business grew. One for customers and billing. One for network monitoring. Another for inventory management. Dozens of incidental spreadsheets that quietly became business-critical. Over time, the real integration lived in people’s heads. There was always someone who knew which system to check, which data was right, and what had changed but had not been updated everywhere yet. That approach works for a while, but it does not scale. 

Most ISPs care deeply about data accuracy and customer experience. The problem is never a lack of effort, or a lack of priority. The problem is that the burden of reconciliation naturally falls on the people. Humans become the glue between systems. When that happens, a few things show up quickly. Work slows down. Knowledge concentrates in silos. Simple questions take too long to answer. Forecasting becomes opaque because accounting lags behind reality. None of that happens because teams are careless. It happens because the systems were never designed to live in the same world. 

Not all data needs to be real time. But some absolutely does. If a customer is on the phone asking whether they can get service at their address, yesterday’s data is not good enough. If construction status changes or a drop is installed, that needs to show up where the work is happening. The same is true for install readiness and scheduling. These are moments where the business is making promises. When systems are not connected in a meaningful way, teams fall back on best guesses or manual checks. That breeds risk. 

Most modern platforms expose APIs. In today’s world, this is table stakes. The important part is everything that comes after. Mapping logic. Error handling. Retries. Logging. Dealing with version changes. Architecting solutions in a way where things cannot fail quietly. When an ISP builds all of that themselves, their systems are connected, but the risk lives with them. In reality, the integration becomes another system that demands care, attention, and expertise. This is where partnerships show their value. When vendors work together directly, they can absorb much of that complexity. They can align on how data behaves, how edge cases are handled, and how changes on one side are reflected on the other without forcing operators to constantly intervene. 

It’s important to remember that the goal of automation is not necessarily to take people out of the process entirely. There are plenty of intersections where a human circuit breaker can be essential. Approving when a plan becomes active. Resolving duplicate records. Deciding how to handle exceptions that logic alone cannot define. Good integrations automate the boring, repeatable work and leave judgment where it belongs 

When thinking about integrations, it’s logical to start with the workflows. What happens when a lead comes in. The rules behind qualifying serviceability. What needs to change when construction finishes. Long before they are ae automated, those workflows already exist. The question is whether they are smooth and predictable or fragile and dependent on tribal knowledge. On the webinar, we focused on high-impact areas like serviceability checks, customer handoffs, and operational visibility. These are places where small improvements compound quickly because they happen dozens or hundreds of times a day. You do not need to automate everything at once. You just need to start where it hurts the most. 

ISPs are being asked to do more with leaner teams. Networks are expanding. Expectations are higher. Data quality matters more than ever, especially when AI-driven automation enters the chat. It is not realistic to expect operators to solve all of this alone. What is realistic is vendors showing up as an ecosystem. Agreeing on data ownership. Aligning on shared objects like locations and readiness. Shipping integrations that work out of the box and continue to work as products evolve. That is the kind of partnership I care about. Not because it sounds nice, but because I have lived the alternative. 

If you want to go deeper, the webinar recording walks through real examples and workflows in more detail. And if you are trying to figure out where to start in your own environment, that conversation is worth having early rather than after the pain sets in. Thanks to everyone who joined the session and shared their real-world perspectives. Those conversations are what pushes this industry forward.