Haris OdobasicThe seven RevOps best practices that drive growth are to focus on revenue, GTM cost and profit; build a team structure that drives alignment; plan, but stay flexible; write documentation that lets the work run without you; automate only when it is easy or at scale; trust data while remembering its limits; and communicate the business impact of every project. Each one moves a number a CEO or CFO already tracks, and each matters more now that AI is entering the revenue engine.
Why these practices matter now
For a CEO or CFO, RevOps is worth funding when it changes how much revenue the go-to-market engine produces, what it costs to produce it, and how predictable it is. That pressure has not eased since we first published this list in 2023. Boards now ask the same question about AI. What will it add to revenue, and is the data ready for it? If you are new to the function, start with our guide to what revenue operations is and how it works.
Our own research shows why the basics still decide the outcome. In our RevOps in 2025 survey of 180 professionals in 28 countries, five times more respondents rated their data quality as poor than as excellent. In the client reviews we ran this year, the same gaps came up again and again, including stale pipeline, stage definitions with no entry or exit criteria, processes that live in one person's head, and automation built before the process was stable.
The seven practices below address those gaps. We have kept them as we first wrote them and added the metric each one should move.
Practice 1. Laser focus on revenue, GTM cost and profit
Many businesses, particularly start-ups, are still working through missed targets, cautious investors and pressure to do more with the same team. Now, more than ever, it is crucial for RevOps to maintain a laser focus on initiatives that raise revenue and profit while reducing GTM costs.
In RevOps, one can easily become engulfed in technical details. However, most of the time, these are less important than one might think. Use the Pareto principle to guide your decisions. Find the 20% of the work that produces 80% of the impact, and do that first.
The metric to watch is GTM cost per unit of new revenue, for example CAC payback in months or GTM spend per euro of new ARR. If a project cannot be tied to revenue, cost or margin, it goes to the back of the queue.
Practice 2. Build a team structure that drives alignment
In RevOps, alignment is crucial to drive activities across departments, be that customer journey mapping, tech stack architecture or identifying revenue leakage.
This implies that RevOps needs to be empowered to enact those changes and must be exposed to a wide range of stakeholders. Where RevOps sits in the organisation has a big effect on that; we compare the options in our post on where RevOps should report.
The structural fix we see work most often is a written set of shared definitions for lifecycle stages, revenue types and handoff rules, each with a named owner. When those live in one revenue handbook, disputes are settled by reference instead of by seniority.
Watch conversion and speed at the handoffs, such as MQL-to-SQL conversion, lead response time and the time from closed-won to onboarding kickoff.
Practice 3. Forget your two-year plan. Stay flexible
It is important to have a RevOps roadmap for the future, but maintaining flexibility is key.
Thanks to the advances in AI and productivity tooling, the software development cycle has never been as short as it is now. This acceleration means that more products are being added to the market than ever before, many of which can significantly improve operational efficiency. However, it also means that a new competitor can enter the market at any time and disrupt your business.
This flexibility becomes even more crucial if you have not yet achieved product-market fit. In practice, that means committing most of the team's capacity to a quarterly roadmap, keeping a fixed buffer for change, and reviewing priorities every month.
The number to track is cycle time from an agreed change to that change being live in the CRM or process. Long cycle times are a sign that the plan has become a cage.
Practice 4. Create documentation that works for you
Your goal with documentation should be to reach a level at which you can take two weeks of leave and nobody notices that you are gone.
You might say that this is impossible, but it truly is not. Every process can be documented. Use screen recordings, for example Loom, for complicated tasks and make sure all information is easy to find.
Perfecting this at a business-wide level can make the difference between a great and an exceptional company. Imagine having no more bottlenecks.
AI is now a second reader of your documentation. An agent that learns only from your CRM records learns how work is done, shortcuts included, not how it should be done. Written stage definitions, qualification rules and ownership are what let you hand work to an AI assistant safely.
Measure it with the two-week test. Count the processes that stop when one person is away. That number should go down every quarter.
Practice 5. Automate when it is easy or at scale. Do not do it too early
Automations are remarkable because they help you scale efficiently. Some quick wins are simple to set up, such as automated email logging, notifications based on triggers or call summaries. Set those up now if you have not done so yet. Selling time is scarce. Salesforce's 2024 State of Sales report, based on 5,500 sales professionals in 27 countries, found that non-selling tasks such as administrative work and meeting preparation take 70% of reps' time (Salesforce).
However, it is not always worthwhile to automate for the sake of automation. Evaluate carefully before you automate your contracting process, for instance. Enterprise contracts often require customisation, and a high volume is needed to justify the automation cost.
In such cases, it is more prudent to focus on improving the process first. Then automate once a consistent, recurring pattern and the right volume have been established.
The most common automation failure we see in client reviews is sequencing. Automation is expected to make up for low adoption or a broken process, and inconsistent data then makes it brittle. The order that works is foundations first, then signal capture, then reporting and automation.
Track the hours of manual work removed per month, measured before and after, plus the volume threshold you agreed before you started building.
Practice 6. Trust in data, but do not forget business is a social science, not STEM
Data is crucial in driving our decisions, but at the same time we have to acknowledge its limitations. Business is a social science. Human behaviour is extremely hard to predict. If it were easy, we would all become rich by buying the right stocks.
We can attempt to create a model that is as close to reality as possible, but there will always be room for interpretation. We are big data fans, but we are also realistic about its limitations. The smaller your business is, the more valuable intuition and experience are. Use data with scale.
For forecasting, that means pairing a data-driven method with a regular cadence of human judgement. Our sales and revenue forecasting best practices cover how to set that cadence up.
The metrics to watch are forecast accuracy over time (forecast against actual, per period) and a pipeline hygiene score per rep, which is a weighted check of close dates, next steps, amounts and recent activity. The score tells you how far the data can be trusted before you lean on it.
Practice 7. Communicate as if your life depends on it
You need to be a good communicator to drive change, and this is especially important in revenue operations. A lot of RevOps work happens behind the scenes. Almost nobody sees it, particularly not the leadership.
The best tip we can offer you is to explain every project in one of three ways.
- how the project increases revenue,
- how it reduces costs,
- or how it increases profit.
Investors pay attention to this. In our survey of 24 venture capital and private equity representatives, 35% of PE firms said they expect RevOps reporting at every board meeting.
It also helps to remember that RevOps influences nearly every revenue KPI but owns none of them. Set goals on levers RevOps can move, such as lead response time, forecast accuracy or data quality, and share the outcome with the team that owns it.
Measure the share of RevOps projects with a stated revenue, cost or profit impact, and how many were reported back to leadership after delivery.
What to stop doing
Good practices fail when they are added on top of old habits. These are the ones we ask clients to drop first.
- Measuring RevOps on numbers it does not own. Revenue belongs to sales and churn to customer success. Measure the levers RevOps moves.
- Keeping three versions of the pipeline report. Name one canonical report per metric and delete or clearly label the rest. Otherwise every meeting starts with a reconciliation instead of a decision.
- Buying tools before the process exists. Zylo's 2026 SaaS Management Index found that organisations leave an average of 36% of their SaaS licences unused (Zylo). Our guide to keeping your RevOps tool stack lean shows how to map what you already have.
- Automating a broken workflow. Fix and document it first, then automate.
- Adding fields because someone might want a report one day. Every field should move the deal forward or answer a question someone actually asks.
- Taking requests through direct messages. One intake channel, one request format and one owner for the backlog.
Not sure which of these practices your team is missing? Our CRM and GTM Health Check ranks the gaps in your data, process and tooling, and fixes the first one.
What changes with AI in 2026
AI does not replace these seven practices. It makes skipping them more expensive.
In our RevOps in 2026 survey of 123 practitioners at AI-engaged B2B SaaS companies, 73% have an AI mandate and 38% have the data to support it. Gartner predicts that over 40% of agentic AI projects will be cancelled by the end of 2027 because of escalating costs, unclear business value or inadequate risk controls (Gartner).
Read against the seven practices, the pattern looks like this.
- Revenue focus decides which AI use case is worth funding. If it does not move revenue, cost or margin, it is an experiment.
- Documentation becomes the instructions an AI agent follows. Data shows what people did; documentation says what they should do.
- Selective automation applies to agents too. Start with the recurring, high-volume step.
- Data you can trust is the precondition. An agent working from stale pipeline repeats the mess at machine speed.
Here is a simple test. If you cannot explain a process clearly enough for a new hire to run it, you cannot hand it to an AI yet. If you want to know which of your processes and data are ready, the Health Check is a fixed-scope way to find out.
FAQ
What are RevOps best practices?
RevOps best practices are the operating habits that make a revenue team efficient and predictable, such as focusing on revenue, cost and profit, aligning teams around shared definitions, keeping plans flexible, documenting processes, automating selectively, using data with judgement and communicating impact. The seven above are the version we use with clients.
What are the four pillars of RevOps?
We work with four pillars, namely data and insights, processes, systems and enablement. Data and insights covers reliable GTM data and reporting, processes covers how each team executes, systems covers the tech stack, and enablement covers onboarding, training and adoption.
How do you measure whether RevOps is working?
Measure the levers RevOps controls rather than outcomes it only influences. Good examples are lead response time, forecast accuracy, conversion at each handoff, data quality scores and hours of manual work removed. Revenue and churn still matter, but their owners sit in sales and customer success.
When should RevOps automate a process?
Automate when a task is simple and recurring, such as activity logging or alerts, or when volume is high enough to pay back the build. If a process still changes every month or depends on custom deals, improve and document it first.
How does AI change RevOps best practices?
AI raises the cost of weak foundations. Agents need documented processes, clear definitions and clean data, so the practices on documentation, data and selective automation become prerequisites for any AI project rather than nice-to-haves.
