Your margin math, on your numbers.

Agency margin is a screening-labour problem before it is a technology problem. This page does not tell you what your margin could be — it gives you the model, shows every formula, and lets you set every assumption yourself.

Open the margin model

The agency surface, shipped today

None of this is a roadmap item. These are the surfaces your consultants and your clients touch now.

Frozen submittal packs

Send one client-safe link per shortlist — up to twenty candidates, frozen at the moment you mint it. The client sees the shortlist, not your workspace: no CV text, no internal notes. The link sits behind an unguessable token with the expiry you choose, and you can revoke it.

Up to 20 candidates · 24h / 7d / 30d expiry · revocable

Re-freeze in place

A shortlist changes after you have sent it. Rebuild the pack under the same URL instead of minting a new one, so the hiring team keeps the link you gave them — and the page tells them when the shortlist last changed.

Same link · rebuilt snapshot · change date shown to the client

Per-open view history

Every open of a pack is recorded with its timestamp, so you know whether the hiring team read the shortlist before the call. We store a truncated user agent and no IP address — the client never agreed to anything here. Read the opens as a signal, not as proof: link previews and crawlers can register one too.

Timestamped opens · truncated user agent · no IP stored

Per-candidate evidence links

Each candidate in the pack carries their own full report link, so a client who challenges a shortlist can read the evidence behind the judgment instead of taking your word for it. Those links survive a re-freeze and expire with the pack.

One report link per candidate · survives re-freeze · expires with the pack

Illustrative — not a customer shortlist.

Senior Backend Engineer · 4 submitted

  • 01 Candidate A — Full report · evidence-cited
  • 02 Candidate B — Full report · evidence-cited
  • 03 Candidate C — Full report · evidence-cited
  • 04 Candidate D — Full report · evidence-cited
Opened 3 times · last opened 2 days ago

The desk system you already run

Workable and Bullhorn write-back are both live: candidate movements and decision summaries land in the ATS your consultants already work in, not in a second system they have to remember to check. Bullhorn is switched on per organisation, so turning it on for your desk is a setting rather than a release. With no ATS, Taali runs standalone on its own job pages with native apply.

Workable · Bullhorn · standalone · public API and webhooks underneath

The margin model

Set every assumption to your own desk. Each result shows the arithmetic that produced it, so you can argue with the model rather than trust it.

Your numbers

Every number here is your input — nothing on this page is a benchmark claim. Change anything and the working below changes with it.

placements

assumption — edit to your desk

£

assumption — edit to your desk

£

assumption — base, commission, employer NI, desk and tooling

hours

assumption — 220 days x 8 hours; edit to your desk

candidates

assumption — edit to your desk

minutes

assumption — CV read, notes and the call-back decision

60%

assumption — your number, not a Taali benchmark. Set it to what you would actually believe.

£

assumption — £0.40 is an example, not a quoted price. Use the rate on your Taali contract.

What that means per placement

Net saving per placement
£242
(£443 − £177) − £24 = £242
Screening capacity multiple
2.50×
12.0 hours ÷ 4.8 hours = 2.50×

Applies to screening time only. It is not a claim about how many roles one recruiter can carry.

Net saving per recruiter, per year
£5,806
£484 × 12 months = £5,806
Recruiter cost per hour
£36.93
£65,000 ÷ 1,760 hours = £36.93
Screening hours per placement — today
12.0 h
60 candidates × 12 min ÷ 60 = 12.0 hours
Screening hours per placement — with Taali
4.8 h
12.0 hours × (100% − 60%) = 4.8 hours
Screening cost per placement — today
£443
12.0 hours × £36.93 = £443
Screening cost per placement — with Taali
£177
4.8 hours × £36.93 = £177
Taali cost per placement
£24
60 candidates × £0.40 = £24
Margin per placement after screening cost — today
£14,557
£15,000 − £443 = £14,557
Margin per placement after screening cost — with Taali
£14,799
£15,000 − £177 − £24 = £14,799
Share of fee left after screening cost — today
97.0%
£14,557 ÷ £15,000 = 97.0%

Screening labour only. Interviewing, client management and business development are not in this number.

Share of fee left after screening cost — with Taali
98.7%
£14,799 ÷ £15,000 = 98.7%

Screening labour only. Interviewing, client management and business development are not in this number.

Net saving per recruiter, per month
£484
£242 × 2.0 placements = £484

Screening labour is the only cost modelled here. Interviewing, client management and business development are not in these figures, so these lines are margin after screening cost, not your whole P&L.

The candidates you already screened

A placed shortlist leaves behind people who were good and simply were not the one. Talent-pool rediscovery re-scores candidates you already have against a new requirement, so a new role starts from your own database rather than from an empty pipeline. The batch is bounded and the estimated cost is shown before you confirm it, and a rediscovery score never overwrites the candidate’s score on their original role.

Bounded batch · cost estimate before you confirm · original scores untouched

Packaging

Per-submittal and per-placement packaging is available alongside standard usage-based pricing. We will price it against your desk economics on a call — talk to us.

Talk to us about packaging

Run the model against a live desk

Bring one role and your own numbers. We will run Taali across the shortlist and you can check the model against what actually happened.