Why Your Intelligence Process Should Be a Line Item, Not a Project
TL;DR: Treating market intelligence as a one-time project means you only synthesize when something has already gone wrong. Growth teams that build it into their operating budget as a recurring function catch signals before they become write-offs. This post explains why the project model fails and what the line-item model looks like in practice.
An intelligence process is the recurring synthesis function that connects what your existing data tools say separately into a single answer to a standing strategic question. Most growth teams don't have one — they find out they needed it after something has already gone wrong.
In October 2023, Convoy — a freight marketplace valued at $3.8B in early 2022 — shut down. CEO Dan Lewis cited "an unprecedented freight market collapse" in a memo to staff. Freight rates had been falling since early 2022 — eighteen months of deteriorating conditions before the closure.
Peloton announced a $400M factory in May 2021 and canceled the project nine months later when demand collapsed and losses mounted.
Also in 2023, 265,660 startup employees were laid off — 59% more than the year prior. This total represents one of the highest single-year concentrations of startup layoffs on record. Companies had plenty of data from their own tools. What they didn't have was a process that connected it.
What Is an Intelligence Process?
An intelligence process is a scheduled, repeatable function that takes data from multiple sources your team already has — analytics platform, CRM, support queue, market data subscriptions — and synthesizes it into a single answer to a standing strategic question. This is the function of a customer intelligence layer.
Dashboards report one source at a time. An intelligence process connects what multiple sources say together and surfaces the relationship between them.
A research project is scoped to answer one question, delivered once, then archived. An intelligence process runs on a cadence — bi-weekly, monthly, quarterly, or triggered by threshold — and is budgeted accordingly.
The distinction matters because it determines what your team actually does when something shifts.
Why the Project Model Fails
Most growth teams commission intelligence as a project: a one-time synthesis effort in response to a specific decision — a pricing review, a GTM pivot, a board ask. The output is delivered, reviewed, acted on, and the function ends.
The problem isn't the quality of that output. It's the gap between outputs.
During the gap, signals accumulate. Retention cohorts shift. Competitive pricing changes. Customer support tickets cluster around a new friction point. None of it gets synthesized because synthesis isn't scheduled.
By the time the next project is commissioned — usually because something has gone visibly wrong — the signal has already become a problem. You're reading the data as a post-mortem, not as an early warning.
Convoy wasn't missing freight rate data. Peloton wasn't missing demand data. The project model meant that data was read only when a decision was already on the table — not as a standing function that would have surfaced the trend earlier.
What a Line Item Looks Like
A growth team is running four data tools: product analytics, CRM, support platform, and a third-party market data subscription. Each tool surfaces data within its domain. Nobody connects them.
The team sees rising churn in their product analytics. The CRM shows deals closing slower. Support tickets mention a competitor's feature three times in the last two weeks. Market data shows pricing pressure in the segment.
Each signal, read alone, suggests a different problem and a different response. Together, they point to a single cause: a competitor made a pricing move that's landing with the segment that drives 40% of revenue. The right response is specific. Only synthesis reveals it.
A team running intelligence as a line item has a process for this. Every month, a synthesis function pulls from all four tools, runs the cross-source analysis, and delivers a structured report: top-line conclusion, ranked findings, action plan, documented gaps. The February report catches the early pricing pressure signal. March is tactical. April isn't a fire drill.
A team running intelligence as a project commissions that synthesis in April, after the churn number appears in the board deck.
How to Budget It
The line-item argument is about timing, not cost. The project model runs behind the pace — commissioned after a need is identified. Cross-source synthesis on a cadence runs ahead of it: signals reach the team six to eight weeks or more earlier than they would under an ad-hoc cadence. For a company with a six-month sales cycle, that's the difference between a proactive pricing adjustment and a retroactive damage assessment.
In running synthesis across growth team data, the timing gap is the first thing that becomes visible — signals that had been accumulating for months arrive in a report after the decision has already been made. The project model doesn't surface them until someone thinks to ask.
A synthesis line item belongs in your budget alongside your tool subscriptions. Without it, the tools produce data that nobody connects.
When a Project Is the Right Model
Not every growth team needs a recurring synthesis function. Three conditions where project-based intelligence is sufficient:
Pre-PMF — if your core strategic questions change week to week, a recurring synthesis run solves for consistency you don't have yet. Build the process after the questions stabilize.
Single data source — cross-source synthesis requires at least two data streams with overlapping subjects. If your decisions rely on one tool, a project delivers exactly what you need.
Hyper specific questions — the line-item model earns its overhead when you already have the data to answer a question. Hyper specific questions typically require unique, qualitative data which you don't already have.
The line-item argument is specifically for teams with standing cross-source questions they haven't been able to answer from a single tool — and who are currently answering them reactively, only when something forces it.
— Steven
FAQ
Should market intelligence be a budget line item or a project?
A project delivers intelligence once, in response to a specific decision, then stops. A line item funds a recurring synthesis function that runs on a cadence — bi-weekly, monthly, quarterly, or triggered by threshold. Teams running intelligence as a project only synthesize after something has gone wrong. Teams running it as a line item catch signals six to eight weeks earlier.
Why does treating intelligence as a project fail?
Because signals accumulate between projects. Retention cohorts shift, competitive pricing changes, and support tickets cluster around friction — and none of it gets synthesized because synthesis isn't scheduled. By the time the next project is commissioned, the signal has become a problem.
What is an intelligence process for a growth team?
A scheduled, repeatable function that takes exports from your existing tools — analytics platform, CRM, support queue, market data subscriptions — and synthesizes them into a structured report against a standing strategic question. Not a new tool. Not a dashboard. A function that connects what your sources say together and runs on a fixed cadence.
What's the minimum viable version of a recurring intelligence function?
One recurring synthesis run against one standing question your team hasn't been able to answer from a single tool, run monthly. That's enough to replace project-triggered synthesis for most strategic decisions.
Scry is the customer intelligence layer for growth teams running cross-source questions on their existing data exports. Learn how it works →