What MCA leads actually cost

01A lead is four different products

The word covers at least four things that share nothing but a name.

• A record. A row of data — business, owner, phone, sometimes revenue. Nobody has spoken to this merchant on your behalf.
• An application. The merchant filled out a funding-request form. The intent is real and it has a date on it.
• A transfer. A call centre reached the merchant, qualified them, and handed the live call to your closer.
• An appointment. A booked time on a calendar.

Price tracks how much human labour happened before the record reached you. It does not track how good the merchant is. A transfer costs more because somebody paid an agent to dial for an hour to produce it — and that agent's hour may or may not be worth more than an hour of your own floor's time. That is the actual question, and almost nobody frames the purchase that way.

02Rough price bands

These are the ranges you will see quoted across the market. Treat them as orientation, not as a benchmark to hold a vendor to.

03Why the spread is so wide

Four variables move the price: how much of each record you get, how old it is, how many other people get it, and how it is delivered. That explains maybe half the range.

The other half is that nobody polices the vocabulary. Fresh means four minutes at one vendor and four months at the next. Exclusive means forever at one and thirty days at another. Two quotes at $25 can be products with a tenfold difference in contact rate, and you cannot tell from the price.

04The number that actually matters

Cost per lead is the wrong unit. Work backwards from a funded deal instead.

Say your average advance is $40,000 and you earn ten points. That is $4,000 of gross commission per funded deal. If you are willing to spend a quarter of that on acquisition, your ceiling is $1,000 in data and labour per funded deal. Now run a bulk file through your funnel and see where you land.

Your rates will differ. Put your own in — the point survives almost any set of numbers you plug in.

Data came to $100 against a $4,000 deal. Two and a half percent. Meanwhile, working 1,000 records to a 20% contact rate takes somewhere between three and five thousand dial attempts, which is roughly a week of one agent's time. At any realistic wage that labour costs ten to twenty times what the data did.

05When paying more is the right call

The answer depends entirely on which resource you have run out of.

If seats are your constraint — three closers, no budget to hire, more hours of demand than you can dial — buy up the ladder. A transfer at $75 that funds one time in eight costs $600 of data per funded deal, which looks terrible next to $100. But it consumes almost none of your dialing capacity. Against a week of agent time, $600 can be the cheaper deal.

If budget is your constraint and you have idle dialer hours, buy down the ladder. You are converting cheap data into deals using capacity you are already paying for.

Most ISOs have this backwards. They buy transfers because the floor is underperforming, when the underperformance is a script and coaching problem that transfers only paper over at ten times the cost.

06Where buyers routinely overpay

• Fresh prices for aged data. Without a date on every row you are taking the vendor's word for the average age of the file. Aged and fresh have no fixed definitions, so this is the easiest place to be quietly overcharged.
• Exclusive prices for a thirty-day window. That is not exclusivity, it is a queue position with an expiry date.
• Columns nobody opens. If your floor only dials, verified email addresses are a line item you are paying for and never using.
• Volume before the funnel is proven. Test at the smallest quantity that gives a readable result, then scale.
• Records you already have. Without deduping against past orders you will buy the same merchant twice and dial them twice, which costs you the data, the agent time, and the merchant's goodwill.

07Six questions that make any quote comparable

1. Does every row carry its own opt-in or filing date?
2. How many buyers receive this record, stated as a number?
3. Over what window, in days, and does the clock reset?
4. Where did the phone numbers come from, and were they checked before delivery?
5. What is the replacement policy for disconnects and bounces, and how long do I have to claim?
6. Will this same record be resold later as aged data?

Six answers in writing turn an unquotable market into a comparable one. Most disputes between buyers and lead vendors trace back to one of these never having been asked.

08Common questions