Most revenue forecasting models fail because they treat agencies like product companies. Taylor Thomson, Head of Finance at performance branding agency WITHIN, thinks that is the root of the problem. His approach integrates qualitative relationship signals with quantitative pipeline data to produce forecasts that reflect the real dynamics of agency work—where deal size, timing, and trust are always in motion.
Traditional forecasting often assumes predictable sales cycles and fixed pricing. In agency environments, however, variables like client relationships, project scope, and budget approvals make prediction far more complex. Thomson’s model acknowledges those nuances, combining human insight with data discipline to create forecasts that executives can actually rely on for strategic planning.
His methodology was instrumental in WITHIN’s growth from $250,000 contract values to $1.8 million enterprise deals. Accurate forecasting gave leadership the confidence to allocate resources, plan expansion, and anticipate revenue flow during the agency’s transformation.
Why Agency Forecasting Is Different
“Agencies don’t sell widgets. We sell relationships,” Thomson often notes. That distinction explains why standard pipeline models break down in service businesses. A single enterprise deal might hinge on months of stakeholder coordination or change dramatically in value as scope evolves.
“We developed the company’s first-ever comprehensive revenue model and dashboard, providing invaluable insights to executive leadership and supporting overall business strategy,” Thomson documented. His model accounts for variables most CRMs ignore: relationship health, stakeholder engagement, and client satisfaction.
The agency model also involves flexible project scopes, negotiable terms, and retention rates tied to relationship quality rather than product demand. For Thomson, the goal was not to impose artificial precision but to capture the probabilities and signals that actually drive business outcomes.
Combining Relationship Signals and Data Discipline
Thomson’s forecasting method blends two elements rarely connected in agency finance: hard data and soft signals. Quantitative metrics such as conversion rates and deal cycle length are paired with qualitative indicators like client sentiment, communication frequency, and engagement quality.
To gather these inputs, his team tracks quarterly satisfaction surveys that achieve response rates above 50 percent, giving an ongoing pulse on relationship health. Early warning signs—such as reduced responsiveness or stakeholder turnover—feed directly into the forecasting dashboard. These qualitative markers often predict retention risks or expansion opportunities before they show up in the pipeline.
This relational layer is what makes Thomson’s system resilient in the face of uncertainty. It reflects an understanding shaped by both his background in political science, where coalition-building depends on trust and consistency, and his years of experience in revenue operations.
Quantitative Rigor That Supports Strategy
Qualitative awareness only works when paired with clean, structured data. Thomson’s insistence on CRM discipline ensures that forecasting models have a solid statistical foundation. “I live and die by Salesforce,” he said in a podcast interview. “If it’s not in Salesforce, it doesn’t exist.”
That rigor allows the team to analyze historical conversion trends, deal timing, and contract variability to identify predictable patterns within unpredictable cycles. The result is not a single-point prediction but a range-based forecast that accounts for uncertainty while remaining useful for resource and hiring decisions.
Thomson’s models also highlight recurring patterns in how enterprise contracts move from opportunity to close. By layering relationship health indicators on top of traditional pipeline data, his team can distinguish between deals that are merely active and those that are truly likely to close.
Accounting for Timing and Contract Variability
Unlike product sales, agency contracts often change size or timeline based on shifting budgets or evolving scopes. Thomson’s system accommodates this reality through probability-weighted scenarios that estimate potential revenue across different confidence levels.
Rather than relying on static close dates, his model uses distribution ranges to reflect potential delays or accelerations driven by client behavior, organizational changes, or budget reviews. The goal is to create forecasts that guide planning without promising false precision.
This adaptive approach has given WITHIN executives visibility into when revenue is likely to land, how much to expect, and which projects carry the highest uncertainty—information that informs staffing, hiring, and cash flow management.
Forecasting as a Leadership Tool
Thomson’s dashboards now serve as a cornerstone of executive planning at WITHIN. Leadership can track revenue trajectory, anticipate team capacity needs, and model different growth scenarios with confidence. The system connects financial forecasts to operational realities, aligning departments around a shared view of performance.
His forecasting framework proved critical during WITHIN’s transition to enterprise-level engagements, where accurate visibility into contract timing and value allowed the agency to scale deliberately rather than reactively.
As Head of Finance, Thomson continues refining the model in collaboration with data science and IT teams, ensuring forecasting evolves alongside AI initiatives and new analytics tools. But the core principle remains the same: forecasting must reflect how relationships, not transactions, drive agency growth.
A Model Built for How Agencies Actually Work
Thomson’s methodology shows that forecasting accuracy improves not by adding complexity but by grounding models in reality. By integrating relationship signals with quantitative discipline, he built a system that captures how agency business truly functions.
The result is a practical framework that balances statistical precision with human context, helping leadership allocate resources, plan growth, and navigate uncertainty with confidence.
For Taylor Thomson, the success of a forecast depends on one thing above all: treating relationships as data that matter.
