For years, agency software has been built around the idea that there should be a single system where the important information about an agency lives. Projects, people, budgets, clients, time and financials should all be brought together so the business has one version of the truth.
It is a sensible idea. It is also not how most agencies actually work.
Agencies use different tools because different tools are good at different things. Proposals might live in Google Docs. Delivery might happen in a project management platform. Time and commercial data might sit in an agency management system. Finance has its own systems. Decisions and changes happen in email, Slack and meetings. None of that necessarily represents a failure of software or process. It is simply the reality of how businesses operate.
The problem is what happens when we assume that any one of those systems contains the complete commercial picture.
A proposal tells you what was pitched and promised. An SOW tells you what was contracted and agreed. An agency management system can tell you about resourcing, rates, capacity and high-level time burn. A project management system can show what happened at task level. Financial systems can tell you what the work ultimately made or lost. Meanwhile, the conversations around the project often contain the context that explains why things changed.
Each source is useful. None is the whole story.
Consider a project that finishes with a poor margin. The financial system can tell you that the margin was poor. The project management system can show where the time went. The SOW can tell you what was originally agreed. The communications might reveal that additional requirements emerged during delivery. But understanding why the margin was lost, and what should be done differently next time, requires connecting those pieces.
That distinction matters because agencies have historically treated data and intelligence as roughly the same thing. They are not.
A report can tell you that a particular type of project regularly overruns. It is more useful to understand whether the overruns came from unrealistic estimates, missing deliverables, a particular type of client requirement, an inadequate discovery process or something else entirely. Better still, that understanding should be available when the next similar project is being scoped.
This is where the idea of a single source of truth starts to break down. The information needed to understand a commercial outcome does not necessarily belong in one system because it was never created in one system in the first place.
Trying to solve that by moving everything into another platform is therefore not necessarily the answer. Agencies have already invested in the systems they use, and those systems exist for good reasons. The more interesting opportunity is to make the information across them useful together.
That requires something different from another system of record. It requires a commercial intelligence layer that can understand the relationships between those records.
A new scope can be considered alongside previous scopes and their actual delivery outcomes. A proposed budget can be compared with the time and resources similar projects actually consumed. A deliverable can be checked against what was included in comparable work. A pricing decision can be considered against the agency’s rate structures, margin rules, historical outcomes and relevant industry benchmarks.
None of this requires the underlying systems to disappear. In fact, it works precisely because they remain where they are.
The broader shift here is important beyond agency software. Businesses have spent decades building systems that record what happened. The next generation of software can start to understand what those records mean when considered together.
Agencies already possess an enormous amount of commercial knowledge. It has been created through every proposal, every project, every timesheet, every financial result and every difficult conversation. The problem is not necessarily generating more information. It is making the knowledge contained within that information available when the next decision is being made.
The single source of truth may never exist.
Perhaps it doesn’t need to.
What businesses need is something that can understand the many sources of truth they already have.
Explore more thinking on commercial intelligence in Keeve Insights.