Cheap leads never convert because the only thing your advertising asked for was the form. Optimise a campaign for the lowest cost per lead and the platform will go and find the people most willing to type their number into anything, which is a different group from the people willing to buy. Nobody in that group has heard of you either, so your sales team spends its week introducing the company to people who cannot remember filling in the form. Cost per lead falls. Cost per customer climbs.
This arrives at our door as a media problem about once a fortnight. The ad account looks healthy, the CPL chart points down and the sales team says the leads are rubbish. Both parties are telling the truth. The number went down because the leads got worse, and no amount of retargeting fixes a list of people who were never in the market.
What is the true cost of a cheap lead?
A cost per lead is a price, and a price only means something next to what you got for it. So here is the same RM20,000 of media in two businesses selling the same product at the same price to the same market. The figures are illustrative and the shape of them is the one we see most weeks.
| Business A, cheapest lead | Business B, qualified lead | |
|---|---|---|
| Media spend | RM20,000 | RM20,000 |
| Cost per lead | RM25 | RM90 |
| Leads | 800 | 222 |
| Contact rate | 35% | 78% |
| Real conversations | 280 | 173 |
| Close rate on conversations | 6% | 18% |
| Customers | 17 | 31 |
| Cost per customer | RM1,176 | RM645 |
Business A pays 72% less per lead and 82% more per customer. Every report it runs says it is winning, because every report it runs stops at the line where the winning ends.
Then there is the cost that never reaches a dashboard at all. Chasing 800 leads at ten minutes each of dialling, waiting and dialling again is 133 hours of somebody's month, against 37 hours for Business B. That gap is most of a full-time salary spent on conversations that were never going to happen, and it is why sales teams stop calling the new leads by week three.
If you want the Malaysian benchmarks these numbers sit inside, our guide to Meta ads costs in Malaysia puts a WhatsApp enquiry at roughly RM10 to RM40 and a booked consultation past RM150. Both figures are normal. Only one of them is a business.
Why does optimising for cheap leads make them worse?
Three mechanisms, and they compound.
The first is that the platform gives you exactly what you asked for. You named the form submission as the event worth money, so the algorithm went looking for humans with a history of submitting forms. It is very good at this. It has no opinion about whether those humans have a problem you can solve, because you never gave it a way to know.
The second is friction, or the removal of it. Meta's instant forms let you choose between a version that submits in two taps and a version that adds a review step before the lead is sent. The fast one always wins on cost per lead. It wins by deleting the half-second in which somebody stops and asks themselves whether they want this at all, and that half-second was doing more qualifying work than your entire targeting setup.
The third is that the offer did the acquiring. When the ad leads with a discount, a free quote or a giveaway, the discount is what people responded to, and you have selected for price sensitivity at the top of your funnel. Google's behavioural research with The Behavioural Architects is the clearest read on how expensive that gets. Across a shopping experiment of 310,000 purchase scenarios, a fictional cereal brand nobody had ever seen still took 28% of shopper preference from the established favourite once it was loaded with five-star reviews and 20% extra free. A fictional car insurer took 87% when it was given an advantage across all six behavioural biases at once.
Most people read that as encouragement, and it is. It is also a price list. An unknown brand can win, and what it pays to win is proof and discount, in every ad, every time, forever. A brand people already recognise pays that once.
What does brand change about a lead?
Brand changes the close rate, and the close rate is the one number your ad account cannot see.
The most useful evidence here comes from marketing mix modelling rather than from platform reporting, because modelling can watch the whole business at once. Analytic Partners, which maintains one of the largest commercial ROI databases in the industry, reports that brand marketing outperforms performance marketing 80% of the time, that upper-funnel work is 60% more effective over the long term and only 25% less effective in the short term, and that 30% of paid search results are directly attributable to brand and upper-funnel marketing rather than to search itself. Their blunter finding is the one worth pinning above a desk: last-click attribution overstates the role of clickable activity by 2 to 10 times.
Read that last sentence next to your own reporting. A meaningful share of what your performance campaigns are currently being credited with was caused by something else, and if the something else is not being funded, the performance numbers decay quietly while everyone congratulates the ad account.
Nielsen's modelling points the same direction from a different angle. Their analysis finds that the long-term impact of media can double the impact of media spend, that ongoing marketing accounts for between 10% and 35% of a brand's equity, and that a brand gives up around 2% of future revenue for every quarter it stops advertising. Nothing in that sentence shows up in a cost per lead.
The practical translation for a Malaysian SME is small and specific. When someone has seen you 3 times before they see the ad, they answer the phone, they know roughly what you charge, they ask about scope rather than about price and they stop shopping you against four competitors on WhatsApp. Nobody logs that as a brand result. It arrives in the sales report as a better month.
Is brand just an excuse for spend nobody can measure?
Often, yes. That objection deserves the first word rather than a footnote, because plenty of brand work is unaccountable and most of the agencies selling it will not commit to a number before they start. If a brand campaign changes nothing you can observe in twelve months, it was decoration and you were right to resent it.
So set the test in advance. Four numbers move when brand work is doing its job, and all four are available to you without a data team: branded search volume, the share of enquiries that arrive direct rather than through paid, the contact rate on your leads and the close rate on conversations. Watch them quarterly against a blended cost per customer. Our guide to measuring marketing ROI beyond ROAS sets out how to calculate the blended figure from numbers you already have.
There is also a genuine case where a performance-only approach is correct, and it should be said plainly. If you sell a true commodity, at a price nobody deliberates over, in a category where no customer holds a preference and nobody buys twice, then harvesting existing demand as cheaply as possible is the whole job. That describes very few businesses. It has never described a business that came to us because its leads stopped converting.
How much of your budget should go to brand?
The research answer and the realistic answer are different, and both are useful.
The research answer is at least half. Analytic Partners recommends assigning no more than 50% of budget to performance activity. Les Binet and Peter Field, working from IPA Databank B2B cases between 1998 and 2018 for the LinkedIn B2B Institute, settle on a 50/50 split between brand building and sales activation, alongside two findings that matter more than the ratio: firms whose share of voice exceeds their share of market tend to grow, and the more famous a campaign makes the company, the better the business results.
The realistic answer is that almost every Malaysian SME we open an account for is running somewhere between 0% and 10% brand, and telling that founder to go to 50% next month is advice nobody can act on. Move to 25% first, hold it for two quarters and watch the four numbers above. 25% of your media budget spent on being known is enough to change the contact rate, which is the cheapest conversion improvement available to most businesses. Nobody needs a rebrand to start. If a rebrand is on the table, our guide to rebrand costs in Malaysia covers what that involves.
Five moves that fix a cheap-lead problem this quarter
None of these needs new budget. Do them in order.
Report cost per customer, not cost per lead. Put the sales outcome next to the media number on the same page, weekly. Most cheap-lead problems survive purely because the two numbers live in two different files owned by two different people.
Add the friction back. Switch instant forms to the version with a review step, ask one qualifying question, and accept that volume will drop. You are paying for a conversation, not a row in a spreadsheet.
Say the price in the ad. A price range in the creative is the fastest qualifier ever invented. It costs you the leads that were never going to pay it, which is the point.
Move 25% of the media above the form. Spend it on being recognised by the people who will buy in the next six months rather than on catching the few buying this week. Reach, video, founder-led content, anything that makes the next ad reach a name that already knows you.
Feed the follow-up. The gap between a lead and a customer is usually a sales conversation with nothing behind it. Give the team a case study, a price guide and a reason to call back a second time.
Frequently asked questions
Should I deliberately raise my cost per lead?
Not deliberately. Raise your standard for what counts as a lead and the cost per lead will rise on its own, because you have stopped counting the submissions that were never worth counting. Judge the change on cost per customer over the following six to eight weeks. If that falls while CPL rises, the campaign got better and the report got more honest.
How do I tell a bad lead from a bad follow-up?
Look at the contact rate first. If most leads never answer, the problem is upstream in acquisition. If most answer but almost none progress, the problem is the offer, the price or the sales conversation. If the leads answer and progress but stall at the quote, you have a positioning problem rather than a lead problem, and we covered where that usually starts in why your brand strategy never made it into your ads.
When does brand work show up in lead quality?
Contact rate and enquiry quality tend to move first, usually inside one to two quarters. Branded search and direct enquiries follow over six to twelve months. Nielsen's finding that a brand loses roughly 2% of future revenue per quarter without advertising works in reverse too, which is why consistency beats intensity here.
Is this different for B2B?
The mechanism is identical and the timescale is longer, because the buying committee is bigger and the cycle runs in quarters. Binet and Field's B2B work puts the brand share of budget at roughly half, which is higher than most Malaysian B2B firms will find comfortable. Start with the contact rate and let the evidence make the argument.
Can I do brand work on a RM5,000 monthly budget?
Yes, and it will not look like a brand campaign. At that level it means consistent creative, a recognisable look, a founder who appears in the content and a message that stays the same for a year rather than changing every campaign. Recognition is built by repetition, and repetition is free. Production value is what costs money.
Does this mean lead ads are a bad format?
No. Lead ads are an efficient way to collect an enquiry from somebody who already wants to talk to you. They become a trap when they are the only thing you run, because then the format's efficiency at collecting form-fills quietly becomes your definition of demand.
The bottom line
A cheap lead is not a saving, it is a deferred cost, and the deferral is short. The cost surfaces in the contact rate, the close rate, the hours your sales team spends on people who will never buy, and eventually in a cost per customer that nobody was tracking while the cost per lead looked so good. Put the two numbers on the same page this week. That single change tells you more than another month of optimisation.
If your leads got cheaper and your revenue did not, talk to us and we will work through your actual numbers, whether or not you end up working with us. You can also see how our performance marketing team reports on customers rather than form-fills, or how brand strategy sets the standard those campaigns are held to.