Backed by Y CombinatorThree documents and three sheets that show exactly how much of your team's capacity ad-hoc requests already consume.
Free download · No account needed
What you get
Pack of 6 files: 3 documents, 3 spreadsheets
Marketing Request Intake Process TemplateCapacity Against Plan
Recomputed from Request Register and Effort by Request Type every period.
| Metric | Value |
|---|---|
| Total quarterly capacity (hrs) | |
| Planned campaign hours | |
| Ad-hoc buffer held | |
| Actual ad-hoc hours consumed | |
| Ad-hoc as % of buffer held | |
| Hours displaced from planned work | |
| Gap in FTE-equivalent |
At this quarter's blended rate, absent a headcount change.
Every marketing team can name its biggest recurring requester from memory. Almost none can say what ad-hoc work actually costs in hours, because the five-minute favor never gets logged and the two-hour deck gets remembered as just this once. Interruption research from UC Irvine found that people compensate for being pulled off task by working faster afterward, at a real cost in stress and effort. That is why a quick ask still belongs in a register instead of being waved through as free. This pack logs everything first, then prices it.
On this pack's own worked quarter, a fictional company's six-person marketing team held a 468-hour buffer for ad-hoc work, 20% of its 2,340-hour total capacity. Real requests over the quarter, 268 of them, cost 735.75 hours, 1.57 times the buffer sized to absorb them. The 267.75-hour overrun came out of already-committed campaign work instead, which is roughly two-thirds of a seventh person's quarter. Sales alone drove half of the ad-hoc hours despite well under half the requests, because its asks average five and a half hours of custom reporting rather than a two-minute edit.
Effort by Request Type turns a request type into an hours estimate from real history, so a new ask is priced before anyone starts rather than after. Capacity Against Plan is the artifact that gets an intake process adopted, because a stated overrun in hours and in a fraction of a headcount is harder to dismiss than a complaint about being busy. When the honest fix is another hire, not more declines, price it the way the marketing plan and budget pack prices everything: backward from what the capacity is worth.
Request RegisterSample rows for a fictional company's marketing team.
| ID | Requester | Team | Type | Effort | Score | Decision |
|---|---|---|---|---|---|---|
| 0004 | Priya Nair | Sales | Sales collateral | 4.5h | 10 | Accept |
| 0026 | Tom Whitfield | Exec | Data pull | 5.5h | 11 | Accept |
| 0041 | Casey Lin | CS | Event signage | 6.0h | 6 | Defer |
| 0070 | Sam Okafor | Mixed | Misc. formatting | 2.0h | 4 | Decline |
| 0091 | Jordan Pace | Sales | New landing page | 3.0h | 3 | Decline |
| 0128 | Priya Nair | Sales | Deck refresh | 8.0h | 6 | Accept (reduced, 1.0h) |
268 requests logged over the quarter. Every row carries an effort estimate before it is scored, never after.
Effort by Type| Type | Count | Avg hrs | Total hrs | % of ad-hoc |
|---|---|---|---|---|
| Sales collateral / deck | 46 | 4.5 | 207.0 | 28.1% |
| Ad-hoc copy or email | 88 | 1.5 | 132.0 | 17.9% |
| Website update | 34 | 3.0 | 102.0 | 13.9% |
| Custom data pull | 28 | 5.5 | 154.0 | 20.9% |
| Event / signage | 12 | 6.0 | 72.0 | 9.8% |
| Brand / legal review | 41 | 0.75 | 30.75 | 4.2% |
| Other / misc. | 19 | 2.0 | 38.0 | 5.2% |
| Total | 268 | 735.75 | 100.0% |
Sales generates 65% of requests by count but half the hours, because most of its asks are quick copy edits, not custom builds.
Capacity Against Plan| Metric | Value |
|---|---|
| Total quarterly capacity | 2,340 hrs |
| Planned campaign hours (80%) | 1,872 hrs |
| Ad-hoc buffer held (20%) | 468 hrs |
| Actual ad-hoc hours consumed | 735.75 hrs |
| Ad-hoc as % of buffer held | 157.2% |
| Hours displaced from plan | 267.75 hrs |
| Gap in FTE-equivalent | 0.69 |
The buffer covered 63.6% of real demand. The rest, two-thirds of a seventh person's quarter, came out of committed campaign work.
One place every request goes, SLA tiers that let genuinely quick asks skip the full rubric, and the arrival steps that log an effort estimate before any work starts.
A four-factor scoring rubric, business impact, deadline hardness, plan fit, and effort, with named point bands for accept, defer, and decline instead of a judgment call each time.
The three-part language for saying no or not yet: what is already committed, what accepting this would displace, and the real alternative, with three worked examples.
Every logged request with its requester, type, effort estimate, score, and decision, so a capacity claim can be checked against real rows instead of taken on faith.
Average hours per request type, computed from the register, so a new ask gets priced from history the moment it is logged.
The number that gets an intake process taken seriously: ad-hoc consumption against the buffer, in hours, percent, and FTE-equivalent, recomputed every period.
Open the pack and send your team's headcount and planning capacity, or download the Word documents and CSV sheets and build the model yourself.
Rebuild Request Register and Effort by Request Type from a shared inbox, a Slack history, or a rough reconstruction of the recurring asks and who sends them.
Capacity Against Plan states what ad-hoc work actually cost against the buffer held for it, in hours, percent, and FTE-equivalent, ready for the next leadership conversation.
Score new requests against Prioritization Criteria, and use Decline Note's format the first time a real ask has to be deferred or declined.
Yes, and the download is not cut down. The zip holds all six files in Word and CSV, no signup and no card. Edit with AI is the other branch: send your team's structure and whatever request history you have, and River builds your own Capacity Against Plan number. More packs in the template library.
Word (.docx) for the three documents and CSV (.csv) for the three sheets, in one zip. The sheets open in Excel, Numbers or Google Sheets and the documents open in Word or Pages, so nothing needs converting. Add ?format=pdf if you want to read them rather than fill them in.
The editorial calendar pack plans and sequences the content the team already committed to publishing. This pack handles everything that arrives after that plan is set: the deck, the data pull, the landing page tweak nobody scheduled. Run both; the calendar's throughput number and this pack's buffer are the same capacity, seen from two directions.
Keep the ticketing tool for tracking status. This pack still adds two things a generic ticket queue does not have: a scoring rubric that turns accept, defer, and decline into a number, and a Capacity Against Plan rollup. That rollup totals what the queue actually costs against the buffer held for it.
No. It is a starting split to test, not a rule. Some teams hold less because their work is almost entirely planned campaigns; others hold more because ad-hoc requests are most of what they do. Run the real numbers for a quarter and adjust the split to match, the same way the worked example's 80/20 was a stated assumption, not a finding.
It creates a free account, installs these six files as a private workspace, and opens by asking for your team's size and planning capacity. Then send whatever request history you have, however rough, and it rebuilds Request Register and Effort by Request Type before computing your own Capacity Against Plan number.
Written by
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.
Send your team's size and whatever request history you have. River logs it, scores it, and states the capacity number in hours, percent, and FTE-equivalent.
Edit with AI