All selected work

Case study 01 / Boldly Grow Media

Decision-support systems

Advertising Budget Planning & Pacing System

From repetitive data collection to an operational decision-support system.

Budget control loop
InputBudget + strategy
  1. 01Plan
  2. 02Observe
  3. 03Project
  4. 04Diagnose
  5. 05Act

Forward planning + continuous feedback

01

Problem

Collecting the numbers was consuming the time needed to interpret them.

Across a growing multi-client portfolio, daily spend collection had become repetitive operational work. At scale, manual collection took about one hour per day. The harder problem was connecting those numbers to a plan and deciding what needed to change.

A useful system needed to answer more than “What did we spend?” It needed to connect the original budget, the observed pace, the likely month-end result, and the action required today.

Daily manual collection at scale
~1 hour
Brands/accounts at peak
12–15
02

Constraints

One operating view. Multiple clients, platforms, and changing rates of spend.

  • Planning had to reflect each client’s budget and strategy across channels and platforms.
  • Actual spend needed a consistent daily structure, even as ingestion methods evolved.
  • Projections had to be understandable enough to use in recurring operational meetings.
  • The system needed to support both forward planning and an ongoing feedback/control loop.
03

Analysis

The reporting task was part of a larger control loop.

Before / current state

A collection task

  • Gather spend figures
  • Enter data manually
  • Interpret the numbers in isolation
After / future state

An operating system

  • Standardize observation
  • Compare plan, actual, and projection
  • Translate a variance into a daily target

The important shift was to connect the stages. A plan establishes intent. Actual spend provides evidence. A projection estimates where the current pace leads. Comparing those views makes exceptions visible and gives the team a basis for corrective action.

04

System Model

Plan → observe → project → diagnose → act.

Budget control loop
InputBudget + strategy
  1. 01Plan
  2. 02Observe
  3. 03Project
  4. 04Diagnose
  5. 05Act

Forward planning + continuous feedback

  1. 01
    Planning engine

    Client budget + strategy → channel/platform plan

  2. 02
    Observation engine

    Actual spend data → standardized daily tracker

  3. 03
    Projection

    Spend to date + recent spend rate × remaining days

  4. 04
    Diagnostic layer

    Compare plan vs actual vs projected outcome

  5. 05
    Action layer

    Calculate corrective daily spend requirements

05

Solution

Connect data collection to a decision, and the decision to an action.

The Google Sheets system combined monthly budget planning, actual spend tracking, pacing, projections, and corrective recommendations. Data collection evolved from manual entry toward automated ingestion using sources and tools such as Triple Whale, Supermetrics, Dataslayer, Google Analytics, and Google Apps Script.

Projected spend

actual spend to date + yesterday’s spend × days remaining

Daily increase needed

budget remaining / days left

Total daily spend needed

current/yesterday spend + daily increase needed

Reading the formulas: an incremental increase must use the projected budget gap. Total unspent budget ÷ days left instead gives the total daily target; adding the current rate to that would overstate the target. Both calculations require days left greater than zero.

  1. 01
    Manual reporting
  2. 02
    Automated ingestion
  3. 03
    Standardized reporting
  4. 04
    Budget planning
  5. 05
    Pacing
  6. 06
    Projections
  7. 07
    Exception detection
  8. 08
    Prescribed daily targets
06

Validation

The system became part of the operating rhythm.

The system was used company-wide and in recurring meetings for roughly one to two years. Its planning, observation, and correction layers were used together in real operational discussions across the portfolio. The evidence is sustained operational adoption: a shared structure for interpreting the numbers and deciding on corrections.

07

Outcome

A shared structure for seeing the situation and deciding what to do next.

What began as repetitive reporting developed into a decision-support system: a consistent place to plan, observe, project, diagnose, and prescribe corrective daily targets. The work connected automation to operational judgment, rather than treating data collection as the end goal.

08

Skills Demonstrated

  • Process analysis
  • Data integration
  • Business rules
  • Workflow design
  • Decision support
  • Automation