Unlocking Business Growth with Analytics in Revenue Operations
by Nihan Yami, Data Scientist, SAS Institute
Revenue Operations (RevOps) is a discipline designed to align different business functions to drive sustainable growth. At its core, RevOps breaks down silos between revenue-generating teams by creating shared goals, metrics, and processes. It ensures that everyone, from marketing teams running campaigns to sales reps closing deals and customer success managers driving retention, is working from the same playbook.
Analytics plays a foundational role here. With unified access to internal systems like CRM and ERP, and external signals from market or customer behaviour, and RevOps enables businesses to build a single source of truth. From this centralised data foundation, we can ask and answer critical questions: Which channels generate the most qualified leads? Where are deals getting stuck in the funnel? Which accounts are most likely to churn?
When embedded into the RevOps process, analytics helps influence pricing strategies, customer journeys, and even product roadmaps.
RevOps covers the full customer lifecycle from pre-sales to post-sales:
- Lead Scoring and Prioritisation: Using models to predict which leads are most likely to convert, helping sales focus their efforts more efficiently.
- Forecasting Revenue: Going beyond spreadsheets by using time-series models and regression techniques to generate more reliable forecasts.
- Churn Prediction: Identifying customers at risk of leaving so that customer success teams can take timely action.
- Upsell and Cross-sell Opportunities: Surfacing accounts with hidden growth potential by analysing product usage, support ticket frequency, or contract history.
The beauty of these models is not only in their predictive power but also in their ability to influence real business outcomes.
The Model Factory: Scaling Analytics with Agility
As the number of models grows, so does the complexity. That’s where the concept of a Model Factory comes in. Think of it as the practical application of MLOps principles; a framework that allows organisations to develop, deploy, and maintain models in a repeatable and scalable way.
There are three pillars to our Model Factory approach:
- Feature Store: A centralised library of engineered variables that can be reused across different projects. This improves consistency and dramatically reduces development time.
- AutoML: Automating parts of the model development process to build faster, benchmark different algorithms, and iterate quickly when business needs shift.
- Monitoring and Drift Detection: Continuously tracking model performance post-deployment. If a model’s accuracy starts to decline, say, because of changes in customer behaviour, we’re alerted early and can take corrective action, such as retraining or rebuilding the model.
This structured approach reduces manual overhead, encourages collaboration, and ensures that models are not just built, but adopted and trusted by the business.
Why Communication is Just as Important as Code
As much as I enjoy the process of building models from data wrangling to tuning hyperparameters, I’ve learned that no model is successful unless it resonates with the audience. That means aligning on the business problem from the start, translating the technical results into actionable insights, and making sure the findings are consumable by non-technical stakeholders.
We constantly ask ourselves:
- Did we define the problem clearly?
- Is this model solving the actual business challenge?
- Who will use the results, and how will they apply them?
- Can we explain the outcome to someone without a data science background?
The most impactful models I’ve worked on didn’t have the highest R-squared score,s but they answered real business questions and got deployed in systems where people could act on them. That’s success in analytics.
A Final Word: Stay Curious and Stay Close to the Business
Analytics in RevOps is not about building the most complex model; it’s about building the most useful one. It’s a field that requires both technical fluency and a deep understanding of the business context. And as AI becomes more democratised, we can all play a bigger role in shaping data-driven decisions.
The best advice I can offer anyone working in this space? Stay curious. Focus on the business problem. Build models that matter. And above all, make sure your work drives action, not just insight.
Join Nihan at BA2025 and discover how to make AI not just possible, but practical.
🗓 Tuesday, 16 September 2025 | ⏰ 10:25 AM – 11:10 AM BST
📍 Convene 133 Houndsditch, London EC3A 7DB
🎤 Building a Model Factory: Streamlining Machine Learning for Business Success
