The Prompt
You are a senior B2B paid media analyst responsible for budget pacing across a multi-channel paid program. You manage 7-figure annual budgets across Google Ads, LinkedIn Ads, Meta Ads, and other paid channels. You understand the cost of underspending (lost pipeline) and overspending (wasted budget) and you make reallocation calls mid-period to keep both in check.
INPUTS
I will paste current pacing data for each active campaign or channel below. Required fields per row: campaign name, channel, monthly budget, spend-to-date, days remaining in the period, current CPL, target CPL, and recent leads.
{PASTE_PACING_DATA_HERE}
{OPTIONAL_PASTE_PRIORITY_NOTES_HERE}
(Examples: "Protect LinkedIn Document Ads spend — Q2 pipeline priority," "Google Brand can accept reduced spend." Leave blank if no priority constraints.)
WHAT I NEED FROM YOU
Produce a pacing report and reallocation recommendation in this exact order:
1. Pacing Status by Channel
For each campaign or channel, calculate and display:
- Spend rate: spend-to-date / (days elapsed in period)
- Required rate: (budget - spend-to-date) / days remaining
- Pace status: Underspending (required rate more than 20% above current rate), On Pace, or Overspending (current rate more than 20% above what's needed)
2. Performance Context
For each underspending or overspending channel, comment on whether efficiency supports the pacing call:
- Underspending + CPL below target: opportunity to scale, not a problem
- Underspending + CPL above target: efficiency issue, may not be worth forcing the spend
- Overspending + CPL below target: working channel, may justify net-new budget
- Overspending + CPL above target: cut, reallocate
3. Reallocation Recommendations
3-5 specific recommendations with dollar amounts. Each must include:
- Source: which channel to pull from, and how much
- Destination: which channel to push to, and how much
- Justification: the data point that supports the move
- Risk: what could go wrong if this reallocation is made
Frame recommendations as proposals for review, not directives.
JUDGMENT RULES
- Do not recommend reallocation away from a channel solely because it's overspending. Overspending on an efficient channel is often the right answer.
- Do not recommend reallocation into a channel solely because it's underspending. Forcing spend into an inefficient channel destroys pipeline economics.
- A channel with insufficient daily volume (under 10 leads in the recent period) cannot be reliably evaluated on CPL alone — flag this and recommend caution before making large reallocation calls.
- If priority notes are provided in the inputs, weight them. Some channels are strategically protected even if pacing math suggests otherwise. Respect that without rubber-stamping it — flag if the strategic call conflicts with strong efficiency signal.
- If you don't have enough data to recommend a reallocation, say so. "Insufficient data on LinkedIn lead quality to recommend a reallocation toward it — recommend pulling MQL conversion data first" is better than guessing.
OUTPUT FORMAT
Return as {OUTPUT_FORMAT}.
If "slack": single message with clear sections, bold dollar amounts, and a summary table of recommended moves at the top.
If "markdown": full structure with tables for pacing status and reallocation moves.
Begin.
How to Use It
Run this prompt mid-period — typically Wednesday morning, or on the 15th of a monthly budget cycle. The timing matters: hitting the midpoint gives you enough days remaining to correct course without panicking. Running it on the last day of the month is too late to act. Claude (Sonnet or Opus) handles the pacing math and performance context reliably; GPT-4 class models work but Claude's output is tighter on the risk-versus-reward framing of each reallocation call.
The key prompt mechanic is the separation of pacing status from performance context. The prompt doesn't just flag underspending channels — it distinguishes between "underspending because the channel is inefficient" (don't force the spend) and "underspending because of bid constraints or creative limitations" (fix the constraint). That distinction matters more than the pacing number itself.
Priority notes are the highest-leverage input. If your VP has told you to protect LinkedIn Document Ads spend in Q2 because it's a strategic pipeline initiative, add that context to {OPTIONAL_PASTE_PRIORITY_NOTES_HERE}. The model will respect it — and will flag if the strategic call conflicts with strong efficiency signal, which gives you something concrete to bring back to leadership.
Example Output
Example output coming soon — currently running this prompt against live data and will publish the redacted output once it's ready.
Common Failure Modes
Variations
Two variations of this prompt are worth knowing.
Variation 1: Single-Channel Deep-Dive
Same pacing logic but scoped to a single platform with campaign-level detail. Useful when you manage a large Google or LinkedIn program with 20-plus campaigns and need to reallocate within a channel rather than across channels.
[PROMPT GOES HERE]
Variation 2: End-of-Period Reallocation
A compressed version scoped to the final 3 business days of a budget period. Prioritizes clearing remaining budget on the highest-efficiency channels rather than full optimization analysis. Pairs well with the Exec Summary Generator to report the reallocation decision to leadership.
[PROMPT GOES HERE]
Get one new prompt every Monday.
Plus the system behind it. Free. Built for in-house demand gen managers at B2B SaaS companies.
Subscribe free →Frequently Asked Questions
Does this work with ChatGPT or only Claude?
Both work. Claude produces more rigorous pacing analysis in my experience — it's better at calling out when the data is insufficient to make a confident reallocation recommendation, rather than producing a confident recommendation table built on thin data. GPT-4 tends to fill every row of the analysis even when sample sizes don't support it. For budget decisions, Claude's "I need more data before recommending" signal is more useful than a confident recommendation that overfit to noise.
What's the best cadence to run this prompt?
Mid-period works best — roughly the halfway point of your budget cycle. For weekly budgets, Wednesday morning. For monthly budgets, the 15th. The prompt is less useful at the start of a period (nothing has spent yet) and at the end (too late to correct course). Running it once mid-period is sufficient for most programs; running it twice — at the one-third and two-thirds marks — is right for high-stakes quarters where you're closely managing pipeline.
How specific should the priority notes be?
Be specific. "Protect LinkedIn" is less useful than "Protect LinkedIn Document Ads — Q2 pipeline priority, VP alignment, minimum $20K monthly floor." The more context the model has about why a channel is strategically protected, the better it can frame the trade-off between the protection and the efficiency data. If you have specific dollar floors below which you won't cut regardless of performance, state those explicitly.
Can I use this for a media agency client's budget?
Yes. The prompt is format-agnostic. You'll need to adapt the B2B SaaS CPL framing to match your client's conversion economics, and add their target CPL or CPA to the pacing data so the performance context section has a benchmark to compare against. The structural logic — pacing math plus performance context — applies to any paid program with defined budget periods and efficiency targets.