There’s so much to unpack here, but it’s a huge lesson for startup founders AND for their customers. For background, ServiceTitan announced that they are no longer integrating with Podium and customers (which it seems like there are ~1,000) have 3 weeks to find a replacement. This is the definition of platform risk. Not because one company is necessarily right or wrong, but because it exposes a truth that both founders and customers routinely underprice.... an integration is not a product feature. It is a revocable treaty between two companies whose incentives can (and will) change. There are a few important lessons here: 1/ Platform risk is an incentive problem, not an API problem. Founders worry about uptime, rate limits, documentation, and technical stability. But the largest risk is strategic and has nothing to do with the tech. An integration remains safe only while the two companies are complementary. The moment the partner begins owning a valuable adjacent workflow, more customer data, or the customer relationship itself, it can (and will) become a competitor. This happens all the time, and is exactly what is happening in the Podium / ServiceTitan relationship. 2/ The system of record owns the veto. A partner can have the better product, stronger customer love, thousands of users, and signed annual contracts. But the platform controlling the system of record still controls the access point. In platform markets, the company creating the most value is not always the company with the power to decide whether that value may continue flowing. 3/ Platform dependent ARR is contingent ARR. So many startups I talk to sell customers annual contracts while relying on platform access that can be changed in hours/days/weeks. Long-term obligations built on short-term permission is a nightmare. If a meaningful percentage of your revenue depends on one company’s API, you do not just have partner concentration. You have a giant liability inside your business model. Investors will evaluate platform concentration the same way they evaluate customer concentration. Not all ARR has the same durability. Although you may be able to grow quickly being a platform partner, the risk and valuation aren't worth it. 4/ Marketplace ecosystems follow a predictable arc. A platform recruits partners to fill product gaps. The partners discover demand, educate the market, and build new categories. The platform watches which categories become strategically important. Then it decides whether to continue partnering, acquire, restrict, replace, or build. The ecosystem is often most open before the platform knows which adjacencies are valuable. Amazon is the king of this.. but it's the same playbook over and over. 5/ Customers ultimately inherit the vendor conflict. An “integrated stack” feels modular until one vendor can sever the connection and leave the customer with overlapping contracts, stranded workflows, retraining costs, and a rushed migration. The important procurement question is not merely: “Does it integrate?” It is: “Who can revoke the integration, what breaks if they do, who owns the data, and how quickly can we recover?” At the end of the day it comes down to this.. Founders should build portability before they need it. And customers should buy continuity, not just connectivity. Because when platforms and partners divorce, the customer pays the breakup fee.