"We need better RevOps" is one of the most common things a growth-stage founder says to me, and one of the least useful, because RevOps isn't a single destination, it's four distinct stages, and the fix that helps a company in stage one will do nothing for a company stuck in stage three. Skip a stage, usually by buying tooling meant for a later one, and you end up with an expensive system nobody trusts, layered on top of a foundation that was never actually built. The model below is the one I use to diagnose an engagement in the first week, before I recommend a single tool or process change.
Stage 1: Ad hoc, revenue data lives in people's heads
This is the default state for most companies under $2M ARR, and there's no shame in it, it's simply what happens before anyone had the bandwidth to build a system. Pipeline lives across a founder's inbox, a spreadsheet someone updates inconsistently, and whatever a rep remembers from their last call. There is no shared definition of a qualified lead, no CRM, or a CRM installed but barely used. The tell for stage one isn't the absence of software, it's the absence of a shared source of truth: ask two people on the team what's in the pipeline right now and you'll get two different answers. The right move here is not a sophisticated attribution model, it's the unglamorous basics, one CRM, defined stages, a named owner on every deal. Anything more advanced than that is solving a problem this company doesn't have yet.
Stage 2: Systematized, the CRM exists but marketing and sales still disagree
Stage two companies have a real CRM, real pipeline stages, and reps who actually log activity. What they don't have is agreement between marketing and sales on what counts as a qualified lead. Marketing reports 300 MQLs a month and feels productive. Sales reports 20 opportunities accepted and feels under-served. Both numbers are accurate and the gap between them is the actual problem, not a data quality issue, a definitional one. The tell for stage two is a recurring, unresolved argument in pipeline meetings about whose numbers are right. The fix isn't more reporting, more reporting just documents the disagreement more precisely, it's a joint MQL-to-SQL definition workshop with both functions in the room, an enforced handoff SLA, and one shared report both teams look at instead of two separate ones.
Stage 3: Aligned, the handoff works but attribution still lies
By stage three, marketing and sales share definitions and the handoff SLA holds. The next failure point is attribution: the company can tell you a lead converted, but not reliably which channel, campaign, or piece of content actually produced it, so budget decisions still default to whoever argues loudest in the room rather than to evidence. The tell for stage three is a functioning pipeline with an unconvincing story about where it came from, ROAS numbers that don't reconcile with backend revenue, last-click credit going to whichever channel happens to touch the form last regardless of what actually built the relationship. The fix is a proper attribution chain, UTM discipline, CRM-level source tracking, server-side conversion data, not a bigger reporting dashboard sitting on top of the same broken inputs.
Stage 4: Optimized, forecasting a CFO would actually underwrite
Stage four is rare, and it's the only stage where the word "predictive" is honestly earned. Pipeline stages have historically-calibrated conversion rates, attribution is trustworthy enough to inform budget reallocation without a fight, and revenue forecasts hold up against actuals within a tight margin, tight enough that a CFO builds a board deck around the number without padding it defensively. Getting here isn't a tooling upgrade from stage three, it's 12-18 months of clean, consistent data accumulating under a system that was already solid. Companies that try to buy stage four with software while still operating at stage two definitional chaos end up with beautiful dashboards built on numbers nobody actually trusts, which is worse than no dashboard, because it looks like rigor while hiding the same underlying mess.
The honest diagnostic
Skip the temptation to self-assess as further along than you are, almost every company I meet initially describes itself as stage three and turns out to be stage one or two once we look at the actual data. Ask three questions. Do two people on the team, asked separately, describe the current pipeline the same way? If not, you're stage one regardless of what CRM you own. Does marketing and sales agree, without an argument, on what qualifies as a good lead? If not, you're stage two even with a mature-looking CRM. Can you trace a specific closed deal back through a defensible, non-last-click chain of what actually influenced it? If not, you're stage three, not stage four, no matter how confident the dashboards look. Fix the stage you're actually in before spending on the one you wish you were in.