Portal leads, aggregator leads or your own database: which gives a finance business better customers

Judge a lead source by the cost of a customer won, not by the cost of a name. For a small loan or insurance office, a sensible order to test is: your own database first, because it costs least per conversation; one aggregator for volume you can actually answer; and portal leads only if you reach people faster than the other offices holding the same lead. A single sheet will tell you which holds for you.

What are you actually buying?

A name, a number and a loan or cover amount, usually sold to more than one office. The customer filled a form once, and within minutes several people ring them. By the time you call they may have spoken to two lenders and been quoted a rate by a third.

That does not make bought leads useless. It makes the contest about speed and about the first conversation. The same lead called in two minutes is a different product from the one called after lunch.

Portals and aggregators also differ. Some portals sell one enquiry to many buyers, and some aggregators pass it to a few. Ask each seller in writing how many buyers receive a lead, what counts as a replacement, and what the customer was told when they filled the form. Treat the answers as claims and test them: ask the people you reach whether anyone else rang first.

What does the sheet look like?

One row per lead, five columns: source, time received, time of your first call attempt, whether the customer spoke to you, and whether a file opened. Add "won" later. Over the same period, work out per source how many you reached, how many became files, and what you spent divided by files won.

Here is the arithmetic for a hundred leads from each source. Every figure is invented for illustration, to show the method and not to suggest what any source delivers.

The portal looked cheapest per name and came out the most expensive per customer. Cost per lead hides that.

The own-database row will look better than it is, because it leaves out the years it took to build the list. Count it as a cheap source you have already paid for.

What went wrong at the Nagpur agency, and how did they change it?

Take an invented two-person loan agency in Nagpur that buys portal leads and calls them between meetings. Their sheet shows the first attempt landing about forty minutes after a lead arrives, and customers picking up on 8 of every 40 calls. They are about to drop the portal.

The sheet points elsewhere. The handful of leads called inside five minutes were answered 14 times out of 20, so the problem looks like their own speed. They set a rule: whoever is free rings within five minutes, and if the call is missed, a WhatsApp message goes out at once saying who they are and offering a time. Only then do they compare the portal with the other two sources, for a fixed number of leads each.

The rule has a price. Whoever is free is often in a customer meeting, and twice in the first fortnight a partner takes a lead call mid-meeting and loses the thread of a file already open. The sample of 20 is also small, so the pattern is a pointer, not proof.

Do not judge a source until you have fixed your own reply time, or you are measuring yourself.

When does your own database win?

When people on it have a reason to be there. A customer whose loan closed, a policyholder whose cover renews, a client who asked a question and got a good answer have all met you. They pick up. They do not compare you with three others in the same hour.

A useful record holds the name, how the person came to you, and what they hold or asked about. It also holds the date of the last contact, the next date something happens (renewal, closure, review), and whether they agreed to be contacted and on which channel.

The catch: an own database is slow. It grows by what you do after each sale, not by spending, and it gives you little if you are opening a new office with no past customers. Then start with one aggregator and build the list from every file you close.

What about consent?

Be conservative. Rules on calls, messages and personal data differ by regulator and change, and a lead sold by a third party may not carry the consent you assume. Ask the seller what the customer agreed to, and in what words. Ask your own adviser or registering body what you may do with the lead and for how long. Keep a record of when and how each person agreed, and never message a bought lead in bulk.

When do you stop paying for a source?

Use two stop rules and one keep rule, and hold your budget and reply time fixed while you judge.

• Stop when you reach customers and still open almost no files. The lead is the problem, not you.
• Stop when the seller cannot say how many buyers receive each lead, or what the customer agreed to.
• Keep when files open but close slowly. That is a follow-up problem, and cutting the source will not fix it.

Run one portal or one aggregator at a time, not three, so the sheet has something clear to say. The wider picture for your trade is on the financial services hub.

Start the sheet with the next lead

Open a sheet today with the five columns and log every lead from now on, whichever source it came from. Choose in advance the number of leads per source you will judge, and do not decide before then. If you would like a second pair of eyes on the sheet, or help with the sourcing side (see demand and lead generation), tell us what you need.