We replaced €15,000 of bought lead lists with AI agents

The best alternative to buying lead lists is to have software find the leads for you, against your own definition of a good client. That's what we built for MYC, a consultancy in Spain that had spent €15,000 on bought lists. Its CRM's agents now find companies that match MYC's own criteria, for cents each, and a person still reads and sends every email.

If you run an agency, a brokerage or a consultancy and you've paid for a list before, this is for you. I'll cover why bought lists tend to disappoint, how the agents work step by step, what a person still has to do, and when this approach isn't worth the money.

Why do bought lead lists so often disappoint?

MYC had spent €15,000 on bought lead lists, delivered as spreadsheets of names.

I don't think MYC was unlucky. A bought list has a few problems built into how it's made.

It's made for everyone, so it fits nobody exactly. A list seller doesn't know what your best client looks like. They sort by broad filters like industry, location and company size, because that's what they can sell to lots of buyers. Your real criteria are usually sharper than that. You want firms of a certain kind, at a certain stage, with a problem you're good at solving.

It starts going stale the day it's made. People change jobs. Companies move, merge, rebrand and close. A list is a snapshot, and you're working from it weeks or months after it was taken.

You may not be the only buyer. Unless the seller says the list is exclusive to you, assume someone else bought it too. That might be a competitor, emailing the same people with a similar offer.

It comes with legal risk you can't hand back. In Australia, the regulator is plain about this. The ACMA's guidance on avoiding spam tells businesses to take care when they buy or use a marketing list, because you are still responsible for having consent for every address you use. Under the Spam Act it's also up to you to prove you have consent. Paying someone else for the list doesn't move that responsibility to them.

And when you need more leads, you pay again.

What's the alternative to buying lead lists?

Build the list yourself, continuously, from your own definition of a good client. Doing that by hand takes hours of searching, checking websites and hunting for the right person to contact. That's the part AI agents are now good at.

In MYC's CRM, the work splits into four steps. We call it Find, Score, Draft, You send.

1. Find. An agent searches for companies that fit MYC's ideal client. The criteria are MYC's, written down once and reused every time.
2. Score. Each company gets ranked against those criteria, with the reasons written out. So the team starts with the strongest matches instead of working down a list from the top.
3. Find the contacts. The agents look for the right people at each company and fill in the record, so nobody is guessing who to write to.
4. Draft. The CRM is connected to MYC's OneDrive. The agents read MYC's own documents (its services, its past work, how it describes itself) and write a first email from that. The drafts are specific to MYC because they're built from MYC's material.

Then it stops. Nothing leaves the CRM until a person has read the draft and pressed send.

The same system reads the inbox, drafts replies and keeps follow-ups on schedule, so a lead found on Monday doesn't get forgotten by Friday. The MYC case study shows the setup.

What does AI lead generation for B2B cost compared with a list?

MYC's lists cost €15,000 in total. The leads its agents find now cost cents each.

There's also the build. An AI CRM like MYC's is a one-off implementation, which starts from A$4,850 with us (or €3,790 in Europe), with no per-seat charge. There's a monthly fee to keep it running, and we put that in writing with the quote. I've broken down the full picture, including off-the-shelf options, in how much an AI CRM costs in Australia.

So you're comparing two different things. A list runs out and you pay again. A system you own keeps finding leads for as long as you run it, and it checks every one against your criteria instead of somebody else's filters.

What does a person still do?

Quite a lot. I want to be clear about this, because "AI lead generation" gets sold as hands-free. It shouldn't be.

First, someone senior has to define the ideal client. The agents are only as good as the criteria you give them, so a partner or owner needs to say who the firm wants to work with, and who it doesn't.

Then someone reads every email. The agents draft, and a person edits each one, sends it or bins it. Your name is on that email, so your judgement should be too.

Someone also has to check the rules. Finding a company doesn't give you permission to market to it. The same consent rules that apply to a bought list apply to leads an agent finds. In Australia that's the Spam Act, and the ACMA guidance linked above is the place to start. Get advice for your situation before you email anyone cold.

And a person has the conversation. Agents can get you to a reply. Turning that reply into a client still takes a phone call, a meeting and a proposal.

What the agents take off your plate is the slow middle bit. Searching, checking websites, ranking, and staring at a blank email.

When is this not worth it?

I'd talk you out of it in three situations.

The first is when you can't describe your ideal client. If the honest answer is "anyone who'll pay", the agents have nothing to aim at. Do that thinking first. It will help your marketing whether you build anything or not.

The second is a tiny market. If there are 40 firms in the country you could ever work with, you probably already know most of them. A spreadsheet and a few good introductions will beat any system.

The third is when your work comes from referrals and you're already at capacity. More leads won't help if you can't take on more clients. Fix the delivery side first.

It tends to pay off when you sell to other businesses, your market is big enough that you can't know everyone, and you've already paid for lists or spent hours prospecting by hand.

What to do this week

1. Add up what you've spent on leads. Bought lists, lead subscriptions, and the hours someone spends searching and copying details into a spreadsheet. Put an actual number on it.
2. Write your ideal client on one page. Industry, size, location, the problem they have, and the signs you'd spot from their website. If you can't fill the page, that's your first job.
3. Check your last bought list. Pick 20 names at random and see how many are still at that company and still relevant to you. That tells you more than any seller's promise.

Paying for leads and not sure they're worth it? Book a 20-minute call and we'll tell you honestly whether a system like this fits, or whether you should keep things as they are. You can also see what's in our AI CRM.