Spend the first 30 days listening: meet every partner and as many founders as you can, and learn where portfolio data lives. Use days 30 to 60 to agree on written goals with the GP and ship one quick win. By day 90, run one repeatable system and report results against those goals.
I relate to this one. Chief of staff roles are famously undefined, and platform roles are much the same. Platform leads tell me their first month felt like getting a title, a laptop and a Slack invite, with the job description to follow. Some are still waiting a year later. One told me the role was "evolving together with me." They meant it kindly, but it also meant nobody had said what good looked like.
That gap is your first real problem to solve. A new hire who waits for the role to be defined gets defined by whoever asks loudest. The ones who do well use the first 90 days to listen, then write the definition themselves and get the GP to agree to it.
Here's a plan in three stages. Adjust the days to your firm, but keep the order.
| Phase | Focus | Output by the end |
|---|---|---|
| Days 1 to 30 | Listen and map | A listening memo and a map of where data lives |
| Days 30 to 60 | Goals and one quick win | A signed-off role charter and one visible result |
| Days 60 to 90 | One repeatable system | A workflow that runs without you pushing it, and a report |
Days 1 to 30: listen, and map where things live
Meet every partner one to one in the first two weeks. Ask each of them the same questions so you can compare answers:
- What should founders say about this firm that they don't say today?
- Which portfolio companies worry you most right now, and why?
- What did the last person in this seat do that you'd want more of, or less of?
- If I could fix only one thing this year, what would you pick?
Then talk to founders, as many as you can, in person where possible. Skip "how can we help?", because founders don't know yet what you can do. Ask two better questions instead. "When did the firm last help you in a way that mattered?" And "When did you last want help and not ask?" The second answer tells you where the platform has been missing.
While you listen, map where portfolio information actually lives. Most new hires find it spread across partner inboxes, board decks in someone's Drive, a CRM nobody updates, WhatsApp threads and a few people's heads. Write it down: what data, where it sits, who owns it, how fresh it is. This map decides what you can and can't do in month three.
End the month with a one-page listening memo. List the themes you heard, the three biggest gaps, and the founder asks nobody followed up on. Send it to the partners. It shows you listened, and it becomes the first draft of your role charter.
Days 30 to 60: agree on goals and ship one quick win
Now write the charter (more on it below) and take it to the GP. Leave that meeting with two or three goals in writing. Platform people who never get explicit goals can't prove their impact later, and that makes the promotion conversation hard. Asking for goals in month two is normal. Asking in year two is awkward.
Good first-year goals are specific and countable:
- Every open role across the portfolio gets candidates from us within a week of posting.
- Every founder ask in a quarterly update gets a response within five business days.
- A running count of customer, investor and talent intros, with what came of them, reported each quarter.
At the same time, ship one quick win. Pick something founders feel, partners notice, and you can finish in two weeks without buying anything. A reliable one: go back through last quarter's founder updates, pull out every ask nobody answered, and answer them. It's unglamorous work, and founders notice right away, because few people read that far into an update.
Days 60 to 90: build one system, not five programs
The temptation in month three is to launch everything at once: a newsletter, a dinner series, office hours, a perks program and a mentor network. Don't. Pick one recurring piece of work and make it run reliably.
The best candidate is usually the work tied to your first goal. If the goal is fast talent help, build the loop: a founder posts a role → you see it the same day → you send two candidates → you log the intro and check back in two weeks. If the goal is answering founder asks, build the capture, so asks from updates and emails land in one list with an owner and a date.
Whatever you build, don't make it depend on people forwarding things to you. They'll do it for two weeks and then stop. Connect to email and calendar where you can, and keep a link from every item back to its source.
If your data map showed portfolio information scattered across inboxes and decks, fixing that early pays off for the whole year. Once the updates, decks and public news for each company land in one place, you see a new role or a buried ask without anyone sending it to you.
Report early, even when it's small
At day 90, send the GP a short written report. Skip the deck. Five lines are enough:
90-day update. Goals we agreed: (1) candidates for every open role within a week, (2) every founder ask answered within five days. So far: 14 roles covered, 11 within a week. 23 asks answered from updates, 4 still open. Working: the role alerts. Not working: asks that need a partner intro are slow. Next quarter: fix that, and start the intro log.
The numbers are illustrative. The format is what matters: the goals, results against them, one honest miss and the next step. A miss builds more trust than a clean sheet, because the GP already knows nothing is perfect.
Then put a review date on the calendar. Priorities tend to go stale after two or three quarters, and the charter should change with them. Your first goals came from the listening memo, and the next ones should come from this report. For help choosing what to count, see what platform teams should measure.
How do new platform hires carve out a role nobody has defined?
Write the definition yourself and get it signed off. After a month of listening, draft a one-page role charter: the three problems you'll own, what you won't do, and how success will be measured. Take it to the GP, agree on two or three goals, and review them every quarter.
A charter can be this short:
- Purpose: one sentence on why the platform exists at this firm.
- I own: three problems, such as talent help for seed companies, founder asks, and the value-add log.
- I don't own (for now): what you're saying no to, such as LP events or deal sourcing.
- Goals: two or three, each one countable.
- Reporting: what you'll share, with whom, and how often.
- Review date: when you'll revisit all of it.
The "I don't own" line is the one people skip, and it's the one that protects your time. Without it, every stray partner request lands on you by default.
What should a new platform hire avoid in the first three months?
Avoid launching big programs before you understand the portfolio, buying tools nobody asked for, building a founder portal, and sending founders surveys or forms. Don't say yes to every partner request either. Each of these creates work that looks busy but rarely moves an outcome the GP cares about.
Here's why each one bites:
- Tools first. A new tool bought before you know the workflow usually becomes one more place nobody looks. Learn the work, then buy.
- A founder portal. Founders rarely log in. Deliver help in email and Slack instead; whether platform teams need a portal goes into the details.
- Surveys and forms. Founders already tell their investors a lot in updates and decks. Read all of it before you ask for anything.
- Programs with no owner after launch. Only start what you can still run in month six.
- Yes to everything. Keep a list of requests you turned down and why. It's useful evidence when you ask for headcount later.
In the first 90 days, you earn trust by being useful to founders in small ways, quickly. The big programs can wait until you know which ones the portfolio needs.