You have been hired because something is not working. The forecast misses, the CRM is not trusted, reps drown in admin, or the tech stack has grown faster than anyone can explain. Nobody says this in the interview. They say "we need more rigour."
Ninety days is the window where you get to ask naive questions and change things without owning the mess you inherited. After that, the problems become yours. This is a plan for spending that window well.
What you'll cover
- Days 1–30: diagnose the revenue engine before touching it
- Days 31–60: fix the forecast and one broken process
- Days 61–90: install an operating rhythm that survives you
- The artifacts you should be able to show by day 90
- The metrics that prove you worked
- Four mistakes that end sales ops tenures early
Why the first 90 days decide the next two years
Sales ops is no longer a reporting function. Gartner expects 75% of the highest-growth companies to run a revenue operations model, and companies with mature RevOps functions are twice as likely to beat revenue goals and 2.3 times as likely to beat profit goals (Gartner).
The catch is scope creep. Gartner also finds that sales and revenue operations teams commonly support five or more groups and spend 68% of their time on non-client work (Gartner). If you do not define your remit in the first 90 days, five other functions will define it for you, one favour at a time.
Days 1–30: diagnose, do not fix
The temptation is to arrive with a plan. Resist it. You cannot see the real constraint from the outside.
Run 15 structured conversations. The CRO, every frontline sales manager, two top reps, two struggling reps, marketing ops, finance, customer success, and enablement. Ask the same four questions each time: What slows you down? What number do you not believe? What did the last person try? What would you fix first? Patterns appear by the eighth conversation.
Audit the pipeline, not the dashboard. Pull the last four quarters of closed deals and open pipeline. Check stage conversion rates, average cycle length by segment, win rate by source, and how often deals slip. You are looking for the gap between the reported story and the data.
Measure data hygiene honestly. Pick 50 open opportunities and score them: next step present, close date in the future, contact roles filled, amount reasonable. Report the percentage. It will be lower than anyone expects, and B2B contact data decays roughly 22.5% a year even when nobody touches it (ZoomInfo), so hygiene is a rhythm, not a project.
Inventory the stack with costs and usage. Every tool, owner, renewal date, seat count, and actual login rate. Gartner's work on sales tech shows most organisations buy far more capability than they use, and roughly 80% of sales chiefs still call the stack critical to hitting target (Gartner). Unused licences are the easiest budget you will ever find.
Deliverable by day 30: a one-page diagnosis. Three findings, one recommended focus, what you are explicitly not doing this year. Present it to the CRO and get it agreed in writing.
Days 31–60: earn credibility with the forecast
Nothing buys you political capital faster than a forecast leadership can rely on. Nothing burns it faster than a reorganisation nobody asked for.
Define the forecast process in one page. Who submits what, by when, using which definitions. Publish stage exit criteria in plain language: what must be true for a deal to sit in stage three. Ambiguous stages are the single largest source of forecast error in most B2B teams.
Introduce one weekly cadence and protect it. A 45-minute pipeline review per manager, same agenda every week: forecast movement, deals that slipped and why, top three deals needing help. No slide decks. The data comes from the CRM or it does not come.
Fix exactly one broken process. Choose the one raised most often in your 15 conversations. Quoting, handoffs from marketing, territory disputes, onboarding new reps. Pick a pilot team, ship the fix, measure it, then roll it out. One completed fix beats four in flight.
Automate the admin the team hates most. Salesforce puts non-selling work at 60% of a rep's time: deck hunting, manual CRM notes, chasing internal approvals (Salesforce). Removing 30 minutes of daily data entry is a more popular win than any dashboard.
Deliverable by day 60: a documented forecast process, a running weekly cadence, and one process improvement with before-and-after numbers.
Days 61–90: build the system, not the heroics
By now the organisation will happily let you become a human help desk. Your job is to make the function scalable instead.
Write the definitions down. A single metrics dictionary: how you calculate win rate, pipeline coverage, cycle length, ramp time, attainment. One definition per metric, one owner, one source. Argue about it once so the company stops arguing weekly.
Stand up one report that replaces ten. A weekly revenue view leadership actually opens. Pipeline coverage by segment, conversion trend, forecast versus commit, risk list. Retire the ad hoc requests it makes redundant.
Set an intake process. Requests come through one queue, are prioritised once a week against your agreed focus, and get a yes, a no, or a date. This is what converts sales ops from reactive to strategic.
Decide the AI question deliberately. Somebody will ask what you are doing about AI. Give a specific answer: one or two workflows where automation removes manual work and every output is reviewable before it reaches a customer. Research summaries, CRM updates from call notes, pipeline hygiene prompts. Avoid anything that writes to the system of record without a human check.
Deliverable by day 90: a 12-month operating plan with three priorities, named owners, quarterly milestones, and the metrics you will be judged on.
Your day-90 evidence pack
If you can show these five things, you have had a good quarter:
- A one-page diagnosis with agreed priorities
- A documented forecast process with published stage criteria
- One process fixed, with measured impact
- A metrics dictionary everyone uses
- A 12-month plan with owners and milestones
Metrics that prove the work
Track a small set and report them weekly from day 30:
- Forecast accuracy: committed versus actual, by segment
- Pipeline coverage: open pipeline against quota, trended
- CRM completeness: percentage of open deals passing your hygiene check
- Selling time: hours per rep per week on customer-facing activity
- Cycle length: median days from qualified to closed
Leading indicators only. If you wait for closed-won revenue to tell you whether the change worked, you have lost a quarter.
Four ways this goes wrong
Reorganising in month one. You will be wrong about the cause, and you will own the disruption.
Building dashboards nobody asked for. Reporting is the output of a fixed process, not a substitute for one.
Saying yes to everything. The 68% non-client time figure is what happens without an intake process.
Skipping the write-up. Undocumented processes revert the moment you take a holiday.
Next steps
Pick one thing from this article and do it this week. If you are early in the role, run the 15 conversations. If you are past day 30, publish your stage exit criteria. Momentum comes from finishing small things in public.
Brazn connects the work across your sales stack so revenue teams spend less time maintaining systems and more time selling. Book a demo to see how it fits your current tools.
