Olivia Chiang, Li Ji, Mateo Perez Barrio, Krisha Singhani, Mikey Taylor, Jaycee Wang
Comparison Table
| Problem 1 | Problem 2 | Problem 3 | |
| Problem | Dependence on other teams’ processes/ capabilities. | Convincing other teams without formal authority. | Conflicting priorities and too many stakeholders (“too many cooks in the kitchen”). |
| Evidence Strength | 6 of 17 interviewees experienced this. | 6 of 17 interviewees experienced this. | 5 of 17 interviewees experienced this. |
| Current Workarounds | Slack follow-ups, recurring status meetings, shared trackers, and spreadsheets. Manual and out of date. | Homegrown solutions, escalating to leadership, and gathering data to build a case. | Slack, Asana, Notion, and a self-built AI agent. Users are trying to solve it themselves. |
| Competitive Gap | Jira, Asana, and Monday track tasks within a team but not cross-team dependencies. | No tool directly addresses influence. It’s mainly handled through meetings and docs. | Project tools assume priorities are already set. |
| ICP | PMs at 200-5k person companies whose work depends on 2+ other teams. | Middle managers at 500+ person companies, who need buy-in from peer teams. | Team leads in cross-functional roles, who juggle requests from many stakeholders. |
| Market Size | SAM: ~$340M. 1.09M US PMs x 25% at 200-5k person firms x 70% with 2+ team dependencies (191k) x 6 seats each (PM + 5 collaborators) = 1.14M seats x $300.
SOM: ~$3-7M ARR in a few years. 1-2% of SAM, or ~11-23k seats across ~2-4k teams. |
SAM: ~$500M. 11.1M US managers x 30% at 500+ person companies (3.3M) x 50% in roles needing peer buy-in (1.7M) x $300.
SOM: ~$5-10M ARR in a few years. 1-2% of SAM, or ~17-33k seats. |
SAM: ~$290M. 11.1M US managers x 35% in cross-functional team-lead roles (3.9M) x 25% at companies already using Slack + Asana (970k) x $300.
SOM: ~$3-6M ARR in a few years. 1-2% of SAM, or ~10-19k seats. |
| Buildability | No: requires deep integrations. | Maybe: scoped tool is possible. | Yes: as a single-user tool using Slack/ Asana data. |
| AI Fit | Maybe: hard to access to other teams’ data. | Maybe-Weak: influence relies on trust. | Strong-Maybe: AI can triage requests & surface conflicts. |
Hardest Part
Market size was the hardest row to fill. There’s no ready-made count of people who face these problems, so we had to estimate how many people match each ICP and what they would pay. Because each estimate depended on our assumptions, the numbers were less reliable than the rest of the table.
Chosen Problem
We choose Problem #3: conflicting priorities and too many stakeholders (“too many cooks in the kitchen”). It’s the only problem we said was fully buildable, because it can start as a single-user tool on top of Slack and Asana without requiring other teams to adopt anything. It also has the strongest AI fit, since AI can triage requests and surface conflicts.
Reading Idea
For developing the ICP, Rachitsky (2023) suggests choosing the three most unique and important traits of the ideal customer from the following list: company size, job title, pain point, way of working, tech used, and business type. We used this to narrow our ICP row, which was originally too broad (e.g. “people who work in companies with more than 1 person”).
