You've found the tool. It cuts your first-pass screening time by hours a week, and you know it'll make your desk faster on the roles that matter. The problem isn't the tool — it's the meeting on Thursday where you have to explain to the managing director why the agency should pay for it. You have a gut feeling it's worth it. Gut feelings don't get budget approved. Numbers do.

This is the gap that kills most tech proposals at agency level: the person doing the screening knows exactly what it costs them in hours and missed placements, but the person signing off on spend only sees a line item. If you want the budget, you need to translate "this saves me time" into "this changes our numbers" — in the language leadership already uses to run the business.

Why the pitch usually fails

Most tech pitches to agency leadership fail for one of three reasons. First, they lead with the tool instead of the problem — "AI screening software" sounds like a cost before anyone's heard the case for it. Second, they use efficiency language instead of commercial language: "saves time" doesn't move a managing director who's thinking in terms of fee income and desk capacity. Third, and most commonly, there's no baseline. If you can't say how long screening currently takes and what it currently costs, you can't credibly say what a tool saves.

Think about what leadership actually tracks: time-to-fill against client SLAs, cost-per-placement, consultant capacity versus live vacancy load, and fall-off rate post-placement. Your pitch needs to speak to at least two or three of these directly, with your own numbers, not vendor claims. A managing director who's watched three pieces of "productivity software" get bought and quietly abandoned will discount anything that sounds like a sales deck. Your job is to sound like you've already done their due diligence for them.

The framework: build the case in four parts

1. Quantify the current cost. Pull real numbers from last month. If a consultant spends 6–7 minutes per CV on a proper screen and you're running 200 applications on a single role, that's over 20 hours of first-pass work before a single qualification call — on one role, while three others sit on the same desk. Multiply that across your live vacancy load and put a number on it: hours per week, per consultant, spent before any client-facing work happens.

2. Tie it to a commercial metric leadership already owns. Time-to-shortlist is the one that matters most, because it's the metric clients notice first and the one most likely to cost you the brief if a competitor agency moves faster. If screening time drops from 20 hours to 3–4 hours on a 200-CV role, that's not a productivity anecdote — it's the difference between a Wednesday shortlist and a Friday one, which affects whether you keep the client relationship at all.

3. Run a paid pilot, not a hypothetical. Leadership doesn't need to trust the concept — they need to see one role prove it. Pick a live, high-volume role and screen it two ways: your consultant's usual process, and a first pass through a tool like CV Matcher to rank and flag candidates against the brief before manual review starts. Track hours spent, shortlist quality, and client feedback on both. A pilot with your own data beats any case study a vendor hands you, and it's low-risk — most tools, including CV Matcher, offer a Start Free Trial that lets you run this comparison before any spend is committed.

4. Frame the upside as capacity, not just savings. The strongest argument isn't "we save 15 hours a week" — it's "those 15 hours let each consultant carry one more live role without adding headcount." That's the line that connects to placement volume and margin, because it reframes the tool from a cost to a capacity increase. If your average consultant carries 8 live roles and could sustainably carry 9 with faster first-pass screening, that's a fee-income argument, not a software argument.

Where you still need the human layer

Be upfront in the pitch about what the tool doesn't do — it makes the case more credible, not less. AI screening ranks CVs against a brief; it doesn't tell you whether a candidate will actually take a counteroffer, whether their reason for leaving is the real one, or whether they'll fit a client's team culture. That still comes from a 20-minute qualification call. No algorithm catches the candidate who looks perfect on paper but goes vague the moment you ask about their actual day-to-day ownership on a project, or the one who undersells themselves in writing but is clearly the strongest technical fit once they're talking through their work. Leadership will respect that you're not pitching a replacement for consultant judgment — you're pitching a way to get to the judgment calls faster, with a cleaner longlist to work from.

What to do before Thursday's meeting

Before you present anything, pull last month's numbers on one high-volume role: CVs received, hours spent screening, and time-to-shortlist. Run that same role's data — or the next comparable one — through a trial screening pass and log the same three numbers. Walk into the meeting with a one-page comparison, not a product pitch. Leadership doesn't need to be sold on AI. They need to see that you've already tested it, it moved the numbers, and you're asking for budget to do more of what's already working.