How much time recruiters lose to bad software

Every vendor in this market will quote you a time saving and none of them know your agency. The honest position is that the loss is real, it is large enough to matter against your margin, and the only number worth acting on is one you measured yourself. That measurement is genuinely a week's work with a tally sheet, and it turns an argument about software into a conversation about money.

What follows is the method we use when an agency asks us to look at how their team works, written so you can run it without us.

Where does the time actually go?

Time is lost in six places, and only one of them feels like a problem while it is happening. Re-entering information that already exists somewhere. Searching for something the system has but will not surface. Waiting for screens, exports or uploads. Moving between tools and getting your bearings again. Correcting data that was wrong or duplicated. And doing by hand a thing the software could do, because nobody knew it could.

The reason none of this shows up anywhere is that each instance is small. Nobody logs forty seconds. A consultant does not go home and say they lost an hour, they say the system is annoying, and the two statements are treated very differently by the person paying the bill.

The largest of the six is almost always re-entry, and the one owners fixate on is almost always waiting. Waiting is visible and irritating. Re-entry is quiet, constant and does not feel like a fault, because typing the candidate's details into the second system feels like the job.

The week-long measurement that gives you a real number

Pick three consultants who are representative rather than your best, give each of them a tally sheet with the six categories on it, and ask them to mark a stroke every time one happens for one week. That is the whole method. It works because a stroke costs nothing to record, and it fails the moment you ask anyone to write down how long something took.

Then time the categories separately rather than in the moment. Sit with one consultant for an hour and time ten real instances of each with a stopwatch: one candidate entered into a second system, one search that went wrong, one export, one correction. You now have a count from the week and a duration from the hour, and multiplying them gives you a defensible weekly figure per head.

Add the switching cost last, because it is the softest. Count how many times an hour a consultant moves between tools, and price each move at the few seconds of navigation plus the longer moment of remembering where they were. A rough version of that number is far better than treating it as zero, which is what every stack review does by default.

Turn hours into the two numbers that change decisions

Convert your weekly hours into money at the loaded cost of a consultant, including national insurance, pension, desk and everything else, rather than at their salary alone. That figure is what the loss costs you to keep, and it is the one to put next to any quote for fixing it.

Then convert the same hours into the activity they would otherwise be. If an hour of a consultant's week is a known number of calls or candidate conversations at your agency, express the loss in those units instead. Owners respond to that in a way they never respond to a cost figure, because it is stated in the currency the business actually runs on.

Both conversions use numbers you already have. Nobody needs an industry benchmark to do this, and an industry benchmark would be worse than your own measurement anyway, since the variation between agencies on exactly this is enormous.

Which losses can you fix without changing system?

Most of them, which is the uncomfortable part. Re-entry between two systems that both stay is fixable with a proper join between them or by dropping one. Failed searches are usually a data problem rather than a search problem, and a fortnight of tidying the fields people actually search on changes the experience more than a new CRM does. Corrections trace back to duplicates, and deduplicating is dull, finite work.

The losses that need a vendor are the ones built into how the software works: slow screens, a texting tool that lives three clicks from the record, an export that exists because a report does not. Those are worth raising as specific requests rather than general complaints, because a specific one occasionally gets built.

Before you start, sort your measured list by hours rather than by irritation. The ranking is almost always different, and the top item is rarely the thing everybody moans about.

Loss	How to measure it	Usual cause	Fix without switching
Re-entry	Tally marks for a week	Two systems that do not join	Join them properly or drop one
Failed search	Count searches that end in a second attempt	Fields left empty, duplicates	Tidy the fields people search on
Waiting	Stopwatch on the ten slowest screens	Heavy list views, big exports	Raise the exact screens with the vendor
Switching	Count moves between tools in an hour	Capability spread across tabs	Move the common actions onto the record
Corrections	Count edits to records made this week	Duplicates and bad imports	Deduplicate once, then police creation
Manual work	Ask what they do more than five times a day	Nobody knew the feature existed	Second training, three months in

What to do with the number once you have it

Give it to three audiences and it does three different jobs. Put it in front of your managers as a target, because a measured hour is something a team can go and win back. Put it in front of your current vendor as a specific list of screens and steps, since that is the only form of feedback that produces changes. And put it next to any switching decision, where it stops the comparison being about licence prices.

Repeat the measurement in six months with the same sheet and the same three people. A number with no follow-up becomes a slide, and the second measurement is the one that tells you whether anything you did worked.

One caution from having watched this go wrong. Do not use the tally sheets to compare consultants with each other, and say so clearly before you start. The moment anyone suspects the exercise is about their performance, the strokes stop being honest and the whole week is wasted.

The losses that never show up on any sheet

There are two costs this method will not catch and both are worth naming. The first is what does not get done because the day filled up, which is nearly always the follow-up call, the update to a candidate who is waiting, and the client conversation that had no deadline on it. The second is what happens to people who spend their week fighting tools, which shows up in your turnover figures rather than anywhere you would look for a software cost.

Neither can be measured with a stopwatch, so I would not try. It is enough to know that whatever number your week produces is the floor rather than the ceiling.

My own opinion is that this exercise is worth running even if you have no intention of changing anything, because it moves the conversation off taste. Software arguments inside agencies are usually two people describing their feelings about a screen. A tally sheet and an hour with a stopwatch turn that into a number, and a number can be argued with, budgeted against and checked again later.

---

Anshika leads Customer Success at Recruitly.