The Staff Who Bring Their Friends

Ask how your best people were hired. A large share came through someone who already worked here.

It is your best channel by a wide margin, and it is almost never asked for on purpose.

Why referrals work

They arrive pre-briefed. Someone has already told them what the work is actually like — the hours, the room, the difficult parts. ⚠️ They have self-selected out of the expectations that cause early leavers.

Somebody vouched for them. Your team member is spending their own standing. They are careful with it, which is a filter no interview provides.

They start with a relationship. Day one is not lonely, which is the single largest predictor of whether the first week goes well.

They stay longer. ⚠️ Partly because leaving means leaving a friend, and partly because they knew what they were joining.

Why nobody asks

It feels like an imposition. Asking someone to recommend a friend feels like asking a favour.

Nobody thinks of it in time. ⚠️ The ask happens when a gap opens, which is the worst moment — the friend is already employed elsewhere and the gap is urgent.

No mechanism. It relies on someone spontaneously mentioning it, which happens rarely and unpredictably.

Asking properly

Ask before you need it. ⚠️ "If you know anyone who would be good here, tell me" as a standing thing, not a request during a crisis.

Be specific about who. "Anyone" produces nothing. "Someone reliable who wants two nights a week" produces names, because it is a description a person can match against people they know.

Say what happens next. They will not recommend someone into a process they cannot describe.

Pay for it. ⚠️ A referral bonus is cheap against agency fees or a bad hire, and paying it in two parts — on starting and on staying — aligns it with what you actually want.

Where it goes wrong

Cliques. ⚠️ The real risk. If half your team came through one person, that person has informal power that nobody granted them, and disagreements become factional.

Uniform hiring. People recommend people like themselves. Over time the team narrows — same background, same view, same blind spots.

The awkward exit. When a referred person does not work out, the referrer takes it personally. Say clearly at the outset that the recommendation is a recommendation, not a responsibility.

Groups leaving together. If a referrer resigns, the people they brought are considerably more likely to follow.

Keeping it healthy

Watch the concentration. ⚠️ If more than a third of the team traces to one person, that is a risk to manage, not a success to celebrate.

Keep one other channel alive. Even when referrals cover your needs. A single-channel pipeline fails completely when it fails, and it brings in a narrower team meanwhile.

Make the standard the same. A referred candidate goes through the same trial as anyone else. Skipping it because they are vouched for is where the bad referral hires come from.

Why referrals work is structural, not sentimental

One  they have already been filtered before you ever meet them
Two  they know somebody on day one
Three they already know what the job actually involves

⚠️ The second is the largest effect and the least discussed. Most leaving happens in the first fortnight, and knowing one person is what carries someone through it.

---

"Let us know if you hear of anyone" is not asking

No time  No number  No role

⚠️ A request with no shape is the same as no request, and it produces the same result. 　 ⇒ "One person who can work Saturdays, starting next month" puts a face 　 　 in somebody's mind. The vague version does not.

Do not announce it to everyone

⇒ ⚠️ Addressed to everyone, it belongs to nobody. Three people asked by name produces more than a notice seen by twenty.

---

Three ways it goes wrong

You do not hire them  and the person who referred them is embarrassed
They leave together  two gaps in one week
Everyone ends up from the same circle

⚠️ The first is the most common and the most preventable. 　 ⇒ Say it in advance: "It may not be right, and if so it will never be 　 　 held against you." Said before, it costs nothing. Said afterwards, 　 　 it is an apology.

---

The second one genuinely happens

Both on the same nights  and the rota depends on the pair

⚠️ Do not let a friendship become a scheduling dependency. 　 ⇒ Split them across nights early, while it is a preference rather than 　 　 a correction.

---

Reward continuation, not introduction

Paid on arrival  you get referrals nobody screened
Paid at three months the referrer screens them for you

⚠️ Move the reward to the far end and the quality moves with it, because the person referring now has a reason to be honest with their friend about the job before they arrive.

---

---

Three to hold

Share of current staff hired by referral. ⚠️ And by which referrer — the concentration is the number that matters.

Retention at six months, referral against other channels. This is the number that justifies paying for it, and it is usually decisive.

Referrals received per quarter. If it is zero, nobody has been asked.

Where the record has to sit

Knowing who referred whom, years later, requires it to have been written down at the time. Without it, the concentration risk is invisible and the retention comparison cannot be made.

tasteck keeps staff records alongside shift history, so how someone came to be here can sit with the rest of their record.

Your best hiring channel already exists. Nobody has ever asked it for anything.

---

Read next

• The Cost Of A Bad Hire In A Small Team
• The Interview That Tells You Nothing
• Three Weeks Into A New Hire
• The Advert That Attracts The Wrong People
• The Advert You Never Took Down