A Guide to Resource Planning for Small IT Agencies

Resource planning is how small IT agencies stop getting surprised (and start operating like pros)#

If you run a small shop, resource planning gives you more control over delivery quality, margins, and cash flow when reality hits. The goal is to replace "tribal knowledge + a calendar" with a lightweight system you can run while juggling Sales, Delivery, and Ops.

Resource planning is deciding what resources you need, when you need them, and how you will use them. In agency workflow terms, you translate pipeline and active work into specific people-hours. Then you make explicit tradeoffs before scope shifts and stalled approvals force them on you.

Here's the operator-level difference:

What this is (and what it is not)#

Resource planning is not a fancy calendar and not a moral judgment about "productivity." It is a decision tool. You use it to answer: Who does what, by when, and what breaks if that assumption changes?

Capacity planning sits alongside it. You aim to keep the right people on the right projects at the right time. This matters most when one senior engineer holds all the context, or when a contractor disappears mid-sprint.

The "business-of-one CEO" approach: simple, reviewable, repeatable#

You do not need heavyweight tooling to start. You need a few consistent artifacts and a weekly rhythm.

Common artifacts include:

• Intake questions that force clarity on scope, timeline, and dependencies before you quote.
• A skills and coverage view (who can do what, and who backs them up).
• A capacity view in hours (work hours, allocated hours, and what remains).
• A short risk list you revisit weekly (approvals, access, waiting on client, vendor lead times).

Tie those artifacts to your Statement of Work (SOW) and to payment reality (invoice timing, deposits, and how long you can float costs) so you protect margins instead of arguing about them later.

And when compliance enters the picture, stay conservative. KYC (Know Your Customer) and AML (Anti-Money Laundering) regulations aim to prevent financial crime, and KYB (Know Your Business) is used to verify business entities. Jurisdictions interpret and enforce these expectations differently. Confirm requirements with your bank, payments provider, or counsel instead of guessing.

Want a quick next step? Browse Gruv tools.

What is IT agency resource planning (and what isn't)?#

IT agency resource planning is the discipline of identifying and allocating the right people (and their time) to client work so you can deliver efficiently and profitably. To run a lightweight system instead of relying on tribal knowledge, you need shared definitions. Most small teams accidentally mix planning, scheduling, and reporting inside a PSA. Then they argue with screenshots instead of fixing the plan.

At the simplest level, treat "agency resource management" as the umbrella term: planning, scheduling, and optimizing how your team is allocated across client projects for efficient delivery and profitability. Resource planning sits inside that umbrella. It focuses on identifying, organizing, and allocating the people, skills, tools, time, and budget needed to complete a project successfully.

Planning vs scheduling (stop faking precision in your PSA)#

Planning answers what should happen and what you will trade off. Scheduling answers exactly when it happens on the calendar. Scheduling is the calendar layer: the when and how of resource usage.

In a 1 to 10 person shop, you do both. Label them explicitly in your PSA (or in Float) so nobody confuses a rough allocation with a locked commitment.

The metrics that keep you honest#

Utilization helps, but it does not run the business by itself.

• Billable hours are hours you charge to a client.
• Non-billable hours keep the business running but do not map to an invoice.
• Billable utilization measures the percentage of available hours spent on client-billable work.
• Throughput is the number of work items you actually finish over a given period.
• Bench time is time when people are not deployed to billable work. Treat it as a signal. Check whether it comes from missing dependencies, slow client inputs, or how work is being sequenced and prioritized.

Guardrail: if you cannot explain your plan in plain language - who does what, roughly when, what triggers Change Control, and how you verify Acceptance Criteria - you do not have a plan you can run. You have vibes.

The small-agency mental model: a 5-gate system you can defend later#

Use a simple decision-gate system so every resource call leaves a paper trail, a rationale, and a clear owner. Once you separate planning from scheduling, you need a repeatable way to say "yes," "not yet," or "no" before you commit scarce capacity.

A gate (tollgate) is a standardized control point where you review a phase and approve or reject moving forward. Even if every opportunity looks "good," you still have limited capacity to execute. Gates help you control intake so you do not commit capacity you do not have.

The 5 gates (operator version) + what to check#

You can name these however you want. Keep the intent consistent and make each gate answer one question.

Your "source of truth" page (one per project)#

Resource management breaks when details are scattered across Slack, inboxes, and someone's memory. Create one shared project page as the single source of truth for the plan, assignments, and progress tracking.

Attach the Statement of Work (SOW) as the anchor document. A SOW defines scope, deliverables, timelines, and responsibilities between parties. That makes it the cleanest reference when the client says, "That wasn't included."

On that same page, keep a lightweight risk register next to the SOW. Keep it short and update it regularly.

• Technical risk: unclear architecture, migration complexity, brittle legacy systems
• Dependency risk: client approvals, access to repos, credentials, procurement delays
• Payment risk: late pay history, long net terms, disputed milestones

You "finish" resource planning when you can answer, without scrambling: (1) who is allocated, (2) what gets deprioritized if reality changes, and (3) which agreement/process governs the change. That is how you run a defensible agency workflow, not vibes.

How do you calculate capacity for a small IT agency without lying to yourself?#

Calculate capacity by converting "hours on the calendar" into "hours you can actually deliver," then validate it regularly against reality. Once you have gates and a source of truth, you need a capacity number you can defend when sales pressure shows up and project allocation gets political.

Start with availability, not hope. Account for vacations, meetings, and focus time. Do not assume all working hours stay usable.

One practical way to run this is a spreadsheet with three lines per person:

• Availability hours (calendar): weekly working hours minus known time off.
• Focus factor (reality check): focus factor = velocity / capacity. In plain terms, it converts "available hours" into "shippable hours."
• Non-project commitments (guaranteed drain): recurring work you will do even if no client project exists.

A concrete starting example: a developer "has 40 hours per week," but with a 0.7 focus factor, "their actual capacity would be 28 hours." If you lack history, pick a starting assumption you can live with, then calibrate it based on recent actual output and interruptions.

The drains you must subtract (use your own list, but make it explicit)#

Do not hand-wave non-project work. List it and subtract it. You do not need an industry-approved list. You need honesty.

Your buckets might include things like:

• Client support and incident work
• Internal operations
• Presales and discovery

If you do not subtract these, you risk "planning" dangerously close to 100% utilization. As professional services guidance puts it: "The fastest way to kill a professional services firm is to push for 100% utilization."

Plan by skill, then probability-weight your pipeline#

Skills-based planning beats headcount planning because it answers the real question: do you have the competencies and knowledge you need, not just "three humans."

Use a simple matrix like this (example lanes):

Finally, forecast demand in hours, not just revenue. A weighted pipeline model applies probability multipliers based on likelihood of closing. Translate each opportunity into "lead hours" weighted by win probability, then staff only what your remaining skill lanes can actually absorb.

What should your weekly/monthly/quarterly cadence look like in a 1-10 person shop?#

Run a lightweight, repeatable cadence that surfaces decisions while they still feel small. Once you can calculate real capacity, your job shifts to protecting it with a routine. The routine should catch allocation risk, pipeline drift, and scope creep before they blow up your delivery dates.

The simplest safe default mirrors how many agencies already operate: weekly for near-term control, monthly for staffing versus demand, and quarterly for strategic re-planning. Keep it boring. Boring scales.

The minimal cadence (and what "good" looks like)#

Use this as your resource planning backbone for resource management, project allocation, and bench management:

A practical agenda template: start with categories that force reality into the open - RAG status, blockers, risks, dependencies, decisions needed, asks, action items with owners and due dates. This structure stops "we talked about it" from masquerading as execution.

Also, do not skip the weekly touchpoint. Do it even if "the team" is just you and two contractors.

Assign owners, even if you wear all the hats#

Even if the same name appears in every row, assign a named owner for each input:

Keep a single change/allocation log as your audit trail. At minimum, capture what changed, the date identified, status, and priority - plus approvals and implementation status. Where relevant, include approval date and a responsible party. If timelines move, record the updated timeline. This habit saves you when a tight SOW turns into a blame spiral.

The operational artifacts competitors skip: templates you can copy today#

If you want resource planning to survive real weeks, not perfect weeks, you need a few boring artifacts that force decisions into writing. Once you've set your cadence, these templates give the process something concrete to run on, even when you feel busy, reactive, and underpowered.

Copy today: the five artifacts that keep your agency workflow honest#

Use this as your starter kit. Keep each artifact to one page or one tab. If it grows, you lost the plot.

Two execution details matter more than the template:

• Make dependencies practical. Don't stop at "waiting on client." Capture what's needed, who owns it, and, when you know it, when it's expected - so dependencies can be managed instead of silently stalling work.
• Track skill-capacity, not just hours. Your rollup should show capacity in the key skill areas you sell, not only total hours, because hours alone fail when one person holds a critical capability.

Definition of Done for a resource plan (so you can defend it later)#

Call your plan "done" when you can point to:

• A completed intake for each active demand item.
• A current allocation view (who does what, when).
• A visible buffer or stated assumption (so "overbooked" shows up as a decision).
• A simple change log for allocation changes (description, date identified, status, priority).

That set of artifacts turns planning into a repeatable system, not a memory test.

Contractors and cross-border realities: how to plan so you can pay people on time (and sleep at night)#

Contractor-heavy planning only works if you plan for onboarding lead time and cash timing, not just "available hours." With templates and cadence in place, this keeps staffing and payouts from turning into a weekly fire drill.

Treat contractors like a mini supply chain (lead time + terms + ramp)#

Contractor resourcing changes the math because you do not control two critical constraints: when someone can truly start and when money becomes safely spendable.

Document these items against the Statement of Work (SOW) start date:

If you negotiated a rate lock for a defined period or made a minimum commitment, write it down. Otherwise margin can evaporate when scope shifts and the contractor rightfully renegotiates.

Build a payout-readiness gate (and minimize PII like a pro)#

Before you promise dates to a client, run a "payout readiness" checklist. This sounds like admin. In practice, it protects productivity and bench management.

Payout readiness checklist (safe defaults):

• Payee details verified (where required by your provider, country, or program). Do not assume one process works everywhere. KYC, KYB (who you do business with), and AML expectations vary by jurisdiction and program.
• Payout method selected and tested: primary method, plus a documented backup if a payout fails or returns.
• Data minimization: PII is personal data. Capture what you truly need, store it only in the right system, and avoid pasting sensitive details into docs and chat. Use masked views in internal trackers (for example, last 4 digits, not full identifiers).

Classification risk stays jurisdiction-specific. Do not "template your way" into misclassification. A local accountant or lawyer can give you a safe confirmation path. Keep a written onboarding checklist per country so your agency workflow stays consistent even when the rules do not.

Tie staffing to cash timing explicitly. Track invoice milestones and payment terms, then add real-world timing friction. Cutoff times, weekends, bank holidays, and payout returns/retries can all add business-day delays.

Gruv's operational model fits here: policy gates plus traceability (request → references → ledger entries → exports), plus payout status tracking and reconciliation exports/artifacts (where enabled). You spend less time asking "where is the money?" when delivery is already running hot.

PSA vs resource planning tools: what to use now, and what to upgrade to later#

Use the simplest tool that lets you run a stable weekly allocation cadence tied to your SOW and Change Control, then upgrade once the process holds. After contractor onboarding and payout timing, the next risk sits in tooling. The usual mistake is buying software to "fix" resource management before you can describe your agency workflow.

Consider starting spreadsheet-first if staffing decisions still live in Slack and memory. A spreadsheet can force clarity: capacity, project allocation, and who owns the decision.

Templates are a practical starting point. Use them until you can answer, every week, "what changed, why, and who approved it" without drama.

Know the boundaries so you don't build a Frankenstack#

A PSA (Professional Services Automation) tool typically aims to cover the full service-delivery lifecycle: project planning and resource allocation, time tracking, billing, and performance reporting.

A resource planning tool focuses more narrowly on allocations and capacity. That narrower focus can beat a PSA when you primarily need cleaner bench visibility, scenario planning, or faster reallocation - depending on the product.

Here's a practical decision frame for agency resource planning:

Examples (not prescriptions): Float markets "Resource Management, Planning & Scheduling Software." Runn markets "real-time resource management and forecasting software." Productive calls itself "Agency Management & Professional Services Automation Software." Scoro positions as "Professional services automation software for consultancies, agencies, IT and A&E firms."

Evaluation criteria that keep small agencies safe#

Use decision-grade criteria, not feature bingo:

The "30 minutes a week" resource planning checklist (and your next best step)#

Treat this as a lightweight system for resource planning, not a one-time planning event. With the core questions answered (capacity, utilization, bench management, cadence), the goal is a repeatable ritual. It should turn new work into a clean project allocation decision you can defend later.

Your safe default is two reviews and a visible workflow. First, a weekly allocation review to reconcile what changed (scope, people, deadlines) and make explicit tradeoffs. Second, a monthly pipeline-to-capacity review to compare likely demand against real capacity so you stop "discovering" conflicts mid-sprint.

Capacity stays simple: resource capacity is the amount of work your resources can take on in a given period, and you can measure it in hours, person-days, or FTE.

Keep your workflow visible with clear stages (use whatever labels fit your shop). If work cannot pass a stage, it does not get scheduled. That rule protects productivity and makes bench management intentional instead of accidental.

One-page checklist (paste this into your ops doc)#

Use this as your decision log and your risk check in one place:

Practical next steps (today, not "someday")#

• Build version 1 in a spreadsheet. Start with a tab for demand (projects and milestones) and a tab for supply (hours by person and skill).
• Schedule the recurring review. Block time weekly. Invite only the people who can change the plan.
• Add one "commercial reality" line to every proposal. State what triggers Change Control and list the client dependencies you require before delivery starts.
• If you operate cross-border, document your process. Keep a checklist for the info you collect and the steps you run, and confirm country-specific requirements with appropriate local guidance (rules vary by jurisdiction).

Want to confirm what's supported for your specific country/program? Talk to Gruv.