In a previous life, I worked on Google Ads (will always call it AdWords), then built an adtech company that managed ~$130M in ads for SMBs. Google’s decision to begin migrating Local Services Ads into Google Ads this August is giving me extreme PTSD. There is much more here than product consolidation... The pay per lead model is staying. But management, reporting, and lead workflows are moving into the same system Google uses for the rest of its advertising products. This is another step toward Google owning more of the contractor/customer relationship. Google does not just want to be where contractors buy demand. It wants to be where they manage it. 1/ Google is moving from discovery to workflow. Google used to be simple. A homeowner searched “HVAC repair near me.” Google sent them to a website. And the contractor owned everything that happened after the click... but that is changing. Google can now control the search result, reviews, verification, lead, call, message, booking experience and eventually what happens after it. Every part of the journey Google can satisfy without sending the homeowner elsewhere makes the contractor’s website less important. I used to think Google would buy a company like Wix to help create websites. Instead, it’s replacing the website. 2/ Google controls demand... so it can keep moving downstream. Contractors spend huge amounts generating demand. But that demand increasingly comes from platforms they do not own. Google, Meta, Angi, Yelp, etc. The contractor performs the work, but the platform sits at the top of the funnel and gets to keep asking what else it can own: The lead? The scheduling? The communication? The payment? The financing? Once you own the customer’s starting point, you get endless opportunities to move deeper into the transaction. 3/ Contractors need to understand the difference between lead generation and customer ownership. Getting a lead from Google is great. Building your entire business around Google sending you leads is something different. If Google controls how the lead gets generated, how much it costs, what information you receive and how the customer interacts with you before the job... then you do not fully own that acquisition channel. This is the same platform risk as the drama between ServiceTitan and Podium, just in a different part of the stack. The platform does not have to do anything malicious. It can simply make decisions that are good for its own business. Contractors ultimately have to adapt, or not I guess. At the end of the day, the interesting part is not where contractors click to manage their Local Services Ads. It is that Google continues moving deeper into the transaction between the homeowner and the contractor. Contractors should absolutely take advantage of that distribution. But the goal cannot be to rent the customer forever. Google can own the introduction. The contractor needs to own the relationship.

Where does an ad-serving stack break as formats keep multiplying? Uber Ads started with Sponsored Listings, then grew to Sponsored Items, display, banners, and video. Retrieval, scoring, and bidding logic forked per surface, and with a shared codebase, deployments became coupled. A bad change on one surface could force a rollback of an unrelated one. The turning point was to rebuild ads around capabilities rather than per-surface features. A supply layer normalizes each surface's request into an AdSpec, a marketplace layer picks an auction or a preferred-deal strategy, and demand providers bid. Surface-specific context is split into extensions, and only the providers that need them use them. Traffic is then split into environment-based fleets running the same codebase. Deployments and rollbacks are independent, and if a provider fails, the auction completes with the remaining candidates. "Build capabilities, not features" and "push complexity to the edges" carry over to any multi-surface platform. Note: this is a blog post, and it reports no quantitative results. Title: Designing a Generalized Ads Marketplace Platform #SystemDesign# #AdTech#

I chatted with @ATechGirl from @supabase on her 25 years of experience as a developer, her time as the project lead of Vitess, her work on Multigres, and databases at Supabase in general. No paywall, give it a read.