I opened Harvest’s renewal invoice after Bending Spoons acquired the company and did the math in forty-five seconds flat.
I had run OSO — a London-based Apache Kafka consultancy — on Harvest for eight years. The tool earned that loyalty. Time tracking, retainer billing, the full loop from time entry to client invoice. It was the kind of software you stop thinking about because it never gives you a reason to.
Then Bending Spoons bought Harvest. The price went up.
The number on the renewal page wasn’t catastrophic on its own. What stopped me was the structure underneath it — per seat, every seat. Every contractor I onboarded: another line item. My team ran at 20 people, but agency headcount doesn’t work like that. You flex up with retainer demand. You bring in a contractor to cover a client spike, deliver the work, move on. That’s how retainer-based consultancies survive.
A per-seat billing tool invoices you for every flex.
I had been paying for a model that financially punishes the growth motion agencies depend on. It took an acquisition and a price increase for me to see it clearly.
The model hadn’t changed. That’s the part that should have worried me sooner.
What Changes When the Price Forces You to Look Harder
The model hadn’t changed. What changed was the price — and once you see it clearly, that the tool invoices you for every contractor you flex in, you start looking at everything else differently.
The logical move is to find a PM platform built specifically for agencies. Not a general-purpose task manager with a time-tracking widget bolted on. An actual agency solution, with retainer billing, time budgets, and profitability reporting inside the product. I’ve run this consultancy long enough to know what the spreadsheet stack looks like without it — the Google Sheet tracking retainer burn, the CSV pulled from the time tracker at the end of every month, the manual re-entry into QuickBooks every billing cycle. You know the mess. You know exactly what a proper tool should replace.
Teamwork.com shows up in every search. "AI-Powered PSA Software Built for Profitability." The positioning is precise. The demo is clean — task management, project views, client work organized the way agency work runs. G2 reviews from agency operators. The entry-tier price looks manageable. The feature list includes time tracking, retainers, invoicing.
I understand why agencies sign up. The tool does what it shows in the demo genuinely well. The decision is not unreasonable.
The problem doesn’t live in the demo. It surfaces the moment you configure your second client retainer and the platform tells you that the feature you evaluated the tool for isn’t available on the plan you purchased.
Where Teamwork.com’s Pricing Structure Actually Breaks for Agencies
That cap has a specific price attached to it.
Teamwork.com puts retainer budget management and client invoicing on its Accelerate plan — $24.99 per user per month, five-user minimum. For a three-person agency, that’s a forced upgrade to roughly $1,500 per year, triggered at the exact moment you need the tool to run an engagement, not before you’ve evaluated whether the engagement is even profitable.
I started pulling apart why the failure doesn’t announce itself sooner.
The UI is engineered to look complete at the entry tier. Time entries appear. Project views render. You can log hours against client projects without friction. The visible surface of the tool behaves like an agency billing platform because the things agencies do constantly — log time, organize tasks, view project status — are all available on the entry plan. The things agencies do at the end of the billing cycle — generate a recurring invoice, check retainer burn against a budget cap, pull a per-project profitability report — are the things locked above it.
The reporting failure compounds the billing failure. Time tracking is on the lower tier. Native reporting is not. To see billable utilization, workload data, or per-project profitability inside the platform, you need Premium — $33 per user per month. The official guidance for agencies on lower plans is to export ticket data to a CSV file and build reports manually outside the product.
That sentence deserves a second read. Teamwork.com’s own support team directs agency operators to maintain their reporting infrastructure in a spreadsheet because the platform’s native reporting is reserved for a higher-paying plan tier.
The architecture is consistent and deliberate: the cost of entering time sits on the entry tier, and the cost of converting that time into invoice data or profitability analysis sits on a separate, more expensive one. You pay per seat to log the hours. You pay more per seat to understand whether those hours made the engagement solvent.
Once you see the structure clearly, you know exactly what every agency on the lower tiers is doing every month to close their books — and it isn’t happening inside the platform.
What Closing the Books Actually Looks Like on a Gated Platform
It happens inside a Google Sheet.
I exported the time entries from Teamwork.com — a CSV pull: raw hours per team member per project, no retainer budget context attached. I opened Google Sheets and manually mapped those hours against the retainer budget I’d been tracking in a separate tab, because the platform wouldn’t enforce it for me. I checked the math. I questioned the math. I checked it again, because a billing error on a client retainer is not a rounding problem — it’s a conversation with the client that costs more time than the reconciliation did.
Then I built the utilization report. Not inside the platform — profitability reporting wasn’t available on my tier. In Excel, column by column: billable hours against capacity, per team member, per client, per month. The report a purpose-built agency billing platform should generate in thirty seconds, I rebuilt from scratch every billing cycle.
Then I opened QuickBooks and re-entered the invoice data I’d spent the last hour reconciling. Time tracking in one system. Budget tracking in a spreadsheet. Invoice generation in a third tool. None of them talking to each other unless I stitched Trello, Slack, Harvest, and Google Docs together with a Zapier chain — and anyone who’s run one of those chains knows how gracefully they handle a contractor who logs time in a slightly different format.
On three active retainers, this rebuild consumed multiple hours per week per billing cycle. The data was never wrong enough to stop the invoice going out. It was never clean enough to answer the question that actually mattered: which client engagement is profitable, and which one is quietly burning capacity I couldn’t price correctly.
I could send the invoice. I couldn’t make the strategic call.
The gap between "good enough to bill" and "good enough to decide" is what finally made me write down the five questions any replacement platform has to answer before I hand it a live retainer.
Five Questions Every Agency Should Ask Before Switching PM Tools
The five questions aren’t theoretical. They came directly out of rebuilding three billing cycles by hand and mapping every hour of manual reconciliation to a specific capability the platform claimed to have but didn’t deliver at the tier I was actually on.
1. Does the platform’s first paid tier include retainer billing — not just time tracking?
I ask the sales rep to walk me through configuring a client retainer budget on the exact plan I’m buying. Not the Enterprise demo. Not the tier three levels up. The plan I’m purchasing today. If retainer budget management is gated above the entry plan, I’m not buying an agency billing platform. I’m buying a task manager with a forced upgrade waiting at the second client.
2. Can I generate per-project profitability reports inside the platform without exporting to a spreadsheet?
During the trial, I open the reporting section and look for billable utilization by client, per-project margins, and workload data. If the answer is "export to CSV and build it in Excel," that is the workaround I am trying to escape. I need to see the profitability report rendered inside the product, on the plan I am on, before the trial ends.
3. What is the actual cost of the first plan that includes retainer features — minimum seats included?
A five-seat minimum on the first retainer-capable plan means a three-person shop pays for two phantom seats before logging a single client hour. I calculate the real entry cost as price per seat multiplied by the minimum seat count, not the per-seat rate in the headline.
4. Does the vendor answer configuration questions within 24 hours during the trial?
I send a real support question on day two — something that requires a substantive answer, not a help doc redirect. If the response takes 72 hours, or the sales rep becomes defensive when I press on the pricing structure, I have learned what the post-sale relationship looks like before I have committed a dollar.
5. Can I track billable hours against a project budget and see burn rate without leaving the project view?
If time tracking lives in the platform, retainer budgets live in a spreadsheet, and invoices live in QuickBooks, the platform has not replaced the manual stack — it has just added another system to reconcile. I need the full quote-to-cash loop in one workspace.
Once I have those five answers in writing, the evaluation is binary. The platform either passes or it doesn’t — and most tools built on the same per-seat, per-tier architecture fail on at least three of them before I finish the trial.
How Nifty Holds Up Against All Five Criteria
I ran those five questions against Nifty. It passes — not because the marketing says so, but because I can verify each one before committing to a paid plan.
Retainer billing at the entry tier. Nifty puts native time tracking, project budget enforcement, and billable rate configuration in the same workspace where the work runs. I can set up a client retainer, configure billable rates, and watch burn rate move against the budget from day one — without hitting a feature wall at the second engagement. The feature the platform sells is the feature available at the tier I’m actually on.
Profitability data inside the product. The audit trail from time entry to retainer drawdown stays inside the platform. I am not pulling a CSV and rebuilding a utilization picture in a separate spreadsheet. Burn rate, time entries, and project budgets sit in the same project view — which means I can answer "is this engagement profitable?" without closing the platform and opening Excel.
Pricing that doesn’t punish headcount. Nifty’s Starter tier runs in the range of $5–$7 per user per month. The first Teamwork.com plan that actually unlocks retainer features costs $24.99 per user per month with a five-seat minimum — a $125 monthly floor before a single retainer hour is logged. Nifty’s free-forever plan also supports unlimited users across two projects, which means I can onboard a new client into the workspace before committing to a paid tier at all.
Integration coverage without rebuilding the stack. Nifty connects to 2,000+ integrations via Zapier. The QuickBooks loop and the HubSpot sync I was running through a fragile automation chain don’t require rebuilding from scratch. I am not replacing one stitched-together system with another.
One workspace for the full quote-to-cash loop. Milestone-driven task automation, native time tracking, and project budget enforcement in a single product is the architectural answer to a platform that separates time logging from billing analysis across two different plan tiers.
If your retainer workflow currently lives in a spreadsheet because your PM tool’s first paid tier doesn’t support retainer billing, you are not running an agency billing platform — you are renting a workaround. Start a Nifty trial against a live retainer, not a demo project, and run the five criteria before you commit.
What that workflow looks like in practice — from retainer configuration to the recurring invoice going out — is what the day-to-day actually changes.
What Running Retainers Looks Like After the Switch
The day-to-day change is not dramatic. That’s the point.
I set up the client retainer inside the same workspace where the work runs. Billable rates configured, project budget enforced, burn rate visible against the retainer ceiling from the moment the first time entry is logged. I don’t open a second tab. I don’t maintain a parallel spreadsheet. The number I need to answer "how much of this client’s retainer is consumed?" lives in the project view, not in a Google Sheet I built last month and will need to update again on Friday.
At the end of the billing cycle, I am not exporting a CSV. I am not reconciling that export against a separate budget document. I am not rebuilding a utilization picture from column data that has no connection to the platform where the hours actually lived. The audit trail from time entry to retainer drawdown is inside one system — which means when a client asks why a particular sprint consumed more hours than quoted, I can show them the milestone breakdown without reconstructing the answer from three different exports.
The milestone-driven task automation replaces the Zapier chain I was running to stitch Trello, Slack, and Google Docs into something that functioned like a connected workflow. When a milestone closes, the dependent tasks move. I stop being the integration layer.
The strategic question — which client engagement is actually profitable — stops being a Friday afternoon exercise in spreadsheet archaeology and becomes something I can answer on a Tuesday morning inside the platform where the work is tracked.
Before, I could send the invoice. I couldn’t make the call about whether to renew the retainer at the same rate.
Now I can do both.
The agencies still running on tools where that distinction doesn’t exist are not just behind on tooling — they’re accepting a structural information deficit that compounds with every client they onboard.
Run the Inspection Before the Client Is Already Onboarded
That structural information deficit has a specific moment when it stops being theoretical.
It arrives when a new client retainer is being onboarded and someone — a partner, an ops director, a CFO — asks a question the platform cannot answer without a spreadsheet export. Which engagements are profitable at the current rate? What is the burn rate on retainer three? Can we onboard a contractor this month without the billing infrastructure breaking?
The danger is never ugly software. Ugly software announces itself immediately. The danger is the clean-looking platform that logs time, renders project views, and passes a two-week demo without ever surfacing the tier wall waiting at the first invoice. By the time the wall appears, you have already told the client the tool can handle the engagement.
I ask three questions before any platform touches a live retainer now. Does the first paid tier — not the Enterprise demo, the tier I am purchasing — include retainer budget management and client invoicing? Can I see per-project profitability inside the platform without opening a spreadsheet? And what is the real entry cost, minimum seats included, not the per-seat number in the pricing headline?
A platform that fails any one of those three in the trial will fail you in the billing cycle. The architecture does not improve after you sign.
If you manage client retainers and you are mid-evaluation, or you are about to onboard a new client and you have not verified whether your current tool can run the engagement without a forced upgrade, run the five criteria from this article against Nifty before you commit. Start with a live retainer, not a demo project. The tier wall reveals itself in configuration, not in a walkthrough.
The spreadsheet stack you built to compensate for the last platform is not a workflow. It is evidence that the platform never actually passed the inspection.
Leave a Reply