How we count

The rule

Every figure is a count over records, worked out at the moment the page loads. There is no stored total anybody can forget to update, no rounded up marketing number, and no way for us to put a figure on a page by hand even if we wanted to.

That also means they go down. An advert finishes and the count drops. A total that only ever rises is a total that has stopped being counted.

A figure appears only once it is real

Nothing is shown until it reaches 3. Not because a small number is shameful, but because "1 business used us last year" reads as a business in trouble, and the honest way to handle a number that is not there yet is to say nothing rather than to round it up.

So if a figure is missing from the home page, it is missing because it has not happened yet. It will appear on its own on the day it does. Nobody has to remember, and nobody can hurry it along.

Why we say offers and not placements

Because an offer is what we can actually see. The employer recorded it, and we know because we sent the email that told the candidate. That is a fact with evidence behind it.

Whether the person accepted, turned up on the first day, or was still there three months later, we do not know, and a board that prints placements is quietly claiming all three. We would rather publish the smaller number we can stand behind than the bigger one we cannot.

What we will never put on this site

• A candidate database size. We do not have one and we are not pretending to.
• A satisfaction percentage from a survey nobody can see.
• A number of "partners" or "clients" that counts anybody who ever emailed us.
• A review score we collected ourselves and cannot be checked.
• Any figure at all before it is true.

Fulfilment, and why ours will look worse

Driving agencies publish a fulfilment rate and it is almost always around ninety eight per cent. There is a reason they all land in the same place: the rate is counted over the shifts the agency accepted. Turn down the awkward ones, the ones at four in the morning, the ones in a town you have nobody in, and the shifts you did accept were nearly all filled. The number is true and it answers a question nobody asked.

Ours counts from the moment a client asks. If somebody rings for a driver and we have to say no, that is a miss and it is in the figure. If a client cancels their own shift it is not, because nothing was left unfilled.

So ours will read lower, and it is the one a transport manager actually wants: if I ring these people at six in the morning, what are the odds. It is also rounded down rather than up, so the figure on the page is never better than the truth.

Four tells, if you are reading somebody else's numbers

We read a lot of recruitment sites while building this one. These came up again and again, and they are worth knowing because they work on you whether or not anybody meant them to.

1. The same number appearing twice for different things. One agency reports that 85% of its placements stay beyond two years and that it has an 85% client satisfaction rate. Two unrelated measurements landing on the same figure is what happens when a number is chosen rather than counted.
2. The same claim with a different figure on another page. Another says it has placed 14,000 candidates on its home page and 10,000 on two others, under three different headings. Nobody is lying. Nobody is counting either.
3. A figure that needs an adjective. "100% of our visible Google reviews are five star." One word is carrying that whole sentence.
4. A traffic figure wearing a person's clothes. "427,000 candidates visited our website in the past year." A website visit is a session, not a person: the same phone coming back on Tuesday counts twice, and a good share of it is never a human at all. And a visitor is not a candidate. Three different things have been squeezed into one word, and the one chosen is the one that sounds like people who want a job.
5. Two totals that were never defined. One site puts "5,241+ candidates placed in the last five years" next to "10,000+ roles filled to date", with clients only since 2019. Nowhere does it say whether a role filled and a candidate placed are the same event counted twice or two different things, so the two figures cannot be checked against each other by anybody, including whoever wrote them. The plus sign on a number precise to the single unit is its own small confession: if you know it is 5,241, the plus adds nothing, and if you do not, the 1 was invented.
6. A total with no period. A driving agency publishes "84,616 shifts fulfilled" and "789,933 hours covered". Since when? The page does not say. Elsewhere it says 1,200 shifts a week and twelve years in business, which multiplies out to something close to three quarters of a million shifts, so the two sets of figures are describing different stretches of time and nothing on the page says which. A total without a start date cannot be checked against anything, including the rest of the same page.
7. No denominator. "85% client satisfaction" out of how many? "Saving an average of five days per hire" measured against what? A percentage with nothing under it is a shape, not a measurement.

And in fairness, one that passes. The same driving agency's two big totals reconcile with each other: 84,616 shifts into 789,933 hours is a shade over nine and a quarter hours a shift, which is exactly what a driving shift looks like. Whatever period they cover, those two numbers were counted rather than chosen, and that is more than most of the sites we read managed. It is worth saying, because a page that only ever finds fault is a page nobody should trust either.

And the one that binds us. We publish page views on the sponsorship page, because a sponsor is entitled to know what they would be paying for. They are called page views there and they will never be called candidates here, or visitors, or reach, or anything else that turns a number of requests into a number of people. If you ever find one of our figures wearing better clothes than it earned, that is a fault and we want to be told.

None of this makes those agencies bad at their work, and two of them are plainly good at it. It means the numbers on their websites are decoration, and you should hold ours to the same test. Everything on this site is a count over records, the record is named, and where a figure is too small to mean anything it is not shown at all.

Reviews, and the one word to watch for

A review on Google, Trustpilot or Feefo is worth something, because the platform holds it and we cannot take it down for saying the wrong thing. When we have enough of those to mean anything, they will be here with a link so you can read the bad ones too.

A quote on a website with a first name under it is worth nothing, because there is no way to check it exists.

And watch the qualifiers. A website vendor we looked at reports that "100% of our VISIBLE Google reviews are 5-star". One word is carrying the whole sentence. A figure that needs an adjective to be true is a figure that is not.

One figure we publish that most do not

Each employer's page shows what proportion of applications they actually answered, measured by us, once there are enough to mean anything. It is a record of what they have done rather than a promise about what they will do, and nothing counts as an answer unless an email really went to the applicant. See who is hiring.