Improving Meta's Self-Serve Ads Manager: Segment, Pain Points and Metrics

You are a product manager working on Meta's Ads Manager, the self-serve tool businesses use to create, run and measure advertising campaigns on Facebook and Instagram. How would you improve it? Choose who you are improving it for, find the problems that matter most to them, propose solutions, pick one, and explain how you would know it worked. ```hint Follow the advertiser's path Trace one advertiser from "I want more customers" through setup, launch and reading the results, and look for the step where they are most likely to give up and stop advertising. ``` ### Clarifying Questions - Which advertisers should the improvement serve: small businesses running their own ads, agencies managing many client accounts, or large in-house marketing teams? - Is the goal more ad revenue for Meta now, better results for advertisers, or keeping new advertisers spending over time, and can we trade some short-term spend for long-term retention? - Does the scope cover the whole flow (campaign setup, budgeting, creative, reporting) or only one stage? - Web Ads Manager only, or the mobile app as well? ### What a Strong Answer Covers - A crisp description of what Ads Manager does and who uses it, with a reasoned choice of one advertiser segment. - Pain points drawn from that segment's journey (objective, audience, budget, creative, measurement), prioritized by how often they occur and how much they cost the advertiser. - Several solutions that differ in kind, compared on impact, effort and reach, with awareness of what Meta already automates. - A north-star metric tied to advertiser success, with guardrails for advertiser results, ad quality and the people who see the ads. - Risks, including whether advertisers will trust recommendations that come from the party selling them the ads. ### Follow-up Questions - Your change raises the share of new advertisers who finish setup, but their average cost per result gets worse. What do you do? - How would your solution change for an agency running dozens of client accounts? - Advertisers compete in the same auctions. How would you design the experiment so treated advertisers do not distort results for the control group?