Your cost per lead is the best it has ever been. Your calendar is the emptiest it has ever been. Both of those things are true at the same time, and they are not a coincidence.
Somewhere in your ad account is a campaign optimizing for leads. It is doing that job well. Last month it brought in forty of them at nine dollars each, which is a number you would have been thrilled with two years ago. Then you look at what those forty leads actually turned into and it is two jobs, one of them small.
The usual explanation is that Facebook leads are junk. That is not what happened. What happened is that you told Meta to go and find leads, you never told it which leads, and it did exactly what you asked with a thoroughness that is honestly a little unnerving. It is the same failure that makes cheap message ads fill your inbox and not your calendar: the number you optimized for went up, and the number you care about did not.
Here is the thing worth sitting with. You know that a roof replacement is worth fourteen thousand dollars to you and a gutter clean is worth three hundred. You know that the customer who books a full system install is worth more than the one who books a filter change, and that a commercial enquiry is worth more than a residential one. Meta knows none of that. Every one of those people fired the same event. As far as the delivery system is concerned, they are identical, and when things are identical it goes and gets you the cheap ones. That is not a flaw in the machine. That is the machine doing arithmetic on the only information it was given.
Two Ways to Tell Meta About Money
There are two separate features for closing that gap, they work completely differently, and almost every article on this subject blurs them into one. Getting the distinction straight is most of the battle.
The first is value optimization, which appears in your ad set as a performance goal called maximize value of conversions. This one works by attaching a real dollar figure to the conversion itself. Your pixel or your Conversions API sends not just "purchase" but "purchase, one hundred and forty dollars." Meta then stops trying to get you the most conversions and starts trying to get you the most total value, which means it will happily pay more for the person it thinks will spend three hundred dollars than for the one it thinks will spend thirty.
The second is value rules. This one does not touch your event data at all. Instead you tell Meta that certain kinds of people are worth more to your business, and it bids more aggressively in the auction to reach them. Not "only show my ad to these people." Just "lean harder when you see one."
They sound like two flavors of the same idea. They are not, and the difference that matters most is who is allowed to use them.
The One You Probably Cannot Use
Value optimization has a threshold, and Meta publishes it plainly. To use maximize value of conversions on website conversions you need at least 100 conversion events with at least 5 distinct values over the past 14 days. For predicted lifetime value events it is at least 100 per week with 5 distinct values over 28 days.
Read that again with your own account in mind. A hundred valued conversions in two weeks is roughly seven a day, every day, each one carrying a real dollar amount back to Meta. An established ecommerce store clears that before lunch. A plumbing company doing twenty jobs a month does not clear it in a quarter, and neither does a dental practice, a law firm, a roofing contractor or the large majority of businesses we work with.
This is the part that gets skipped, and skipping it is how an owner burns an afternoon trying to switch on a setting that was never going to be available. If you are running a local service business, value optimization is probably not your tool. Not yet, and possibly not ever.
That is not the end of the article. It is the reason for it.
The One You Can Use Today
Value rules carry no volume requirement at all.
You will find them in Advertising settings, or you can create them directly inside an ad set, and then you apply them per ad set. The criteria Meta lets you build a rule on are age, gender, location, operating system, device platform, and a selection of placements and conversion locations. For each rule you choose a criterion and an adjustment, and Meta will let you go as far as increasing a bid by 1000 percent or decreasing it by 90 percent.
What that adjustment actually does is worth being concrete about, because "increase bid by 50 percent" sounds more dramatic than it is. Every impression on Facebook and Instagram is an auction. Setting a rule does not buy you anything or guarantee you anyone. It tells the delivery system it is allowed to compete harder when the person on the other end matches your rule. You are not selecting an audience. You are raising your hand higher for part of one.
There is one catch that will stop you before you start. Value rules are not available with manual bid strategies. If your campaign is running on ROAS goal or bid cap, the option is not there. You need an automatic strategy, which for most small accounts means highest volume or a cost per result goal, and most small accounts are already there.
Read Your Own Account Before You Set a Rule
The temptation now is to go and set a rule based on who you think your best customer is. Resist it for twenty minutes.
Pull up the last ninety days of a campaign that has produced real business, and break the results down by age, then gender, then placement, then location. You are looking for one thing: a segment where you spent a meaningful amount of money and got a meaningfully better outcome. Not a segment with four conversions and a flattering cost per result, which is noise wearing a suit. A segment with enough spend behind it that the difference would survive another month.
Often there is nothing there, and that is a real finding. The breakdown is flat, every age band performs about the same, and the honest move is to leave value rules alone and go fix something else. A value rule set on a hunch is just an expensive opinion, and because it changes what you pay rather than what you show, it is an opinion that costs you on every single impression.
Sometimes there is something obvious. Desktop converts at four times the rate because you sell to businesses and people research you at work. One metro area produces double the job size of the others because that is where the older housing stock is. Women outspend men by a third. Those are the moments value rules were built for.
And a warning that catches people in exactly the industries that can least afford it: if your campaign runs in a special ad category, which covers housing, employment and credit, then age, gender and certain location criteria are restricted. They may still look selectable while you are building the rule. Choosing one produces an error and gets the ad rejected. If you are a property manager, a recruiter or a lender, build your rules on device, platform or placement and leave the demographics alone.
How Much More Is Actually More
Once you have found a real difference, you have to price it, and the instinct is to reach for a big round number.
Meta's own guidance is better than the instinct, and it is arithmetic rather than vibes: if you earn $115 from women and $100 from men, women are worth about 15 percent more to your business, so you set a 15 percent rule. The adjustment should come out of your own numbers. If you genuinely have nothing to work from, Meta suggests starting at 50 percent and watching what happens, which is a reasonable place to begin and a terrible place to stay.
Then there is the trap, and it is the thing most likely to make a carefully built rule set quietly do nothing. When two rules overlap, only the first applicable one fires. Meta's own example: rule one says bid 20 percent more for women in a particular state, rule two says bid 50 percent more for women using a particular mobile operating system. A woman who is in that state and on that operating system matches both. She gets 20 percent, because rule one is higher in the order. Not 70 percent, not 50 percent. Twenty.
Which means the order of your rules is not cosmetic. It is the logic. Put your highest-value, most specific rule at the top, or the broad one above it will swallow every person the specific one was written for.
Your Cost Per Lead Is Going Up. That Is the Point.
When you build your first rule set, Meta makes you tick a box confirming you understand your cost per result may increase.
It is not covering itself. It is telling you what the feature does. You have just instructed the auction to pay more for certain people. You will pay more for those people. In the language of the only report most owners look at, this setting makes your account look worse.
That is the whole trade, and it only works if you are prepared to stop judging the account on cost per lead and start judging it on what the leads were worth. Every lead generation client any of us has ever worked with would take a lead at double the price if it closed at triple the rate. Almost none of them have an ads dashboard set up to notice when that happens.
The honest version of this also means acknowledging that the trade does not always pay. One practitioner ran value optimization against plain conversion volume as a controlled test on an ecommerce account, a month of data, deliberately outside the Black Friday window. Value optimization did what it promised: average order value came in at about $100 against $88 for the volume campaign. It also cost $61 per purchase against $50. The higher basket did not cover the higher price, and on return on ad spend the boring campaign won.
That result is not an argument against any of this. It is an argument for measuring it properly and for not betting the account on it. Meta's own documentation says the quiet part directly: value rules help the system focus your budget where it is likely to matter most, and they do not guarantee more conversions. Run this on part of your spend, give it long enough to mean something, and judge it on revenue rather than on the number that got worse by design. There is a value rules breakdown at the ad set level that shows results per rule, which is where you go to find out whether the thing you paid extra for showed up.
Meta Took Control Away. Then Handed Some Back.
Step back from the buttons for a moment, because there is a pattern here that tells you where this is going.
For several years the direction of travel has been one way. Detailed targeting became a suggestion. Lookalikes became a suggestion. Age and gender inputs became suggestions. Turning off placements got harder and the option kept moving. We have written about that shift on its own, because your ad is now doing the job your targeting used to do, and broadly it has been good for small advertisers who were never going to out-guess the machine anyway.
Then, in the middle of all that, Meta built value rules and started actively pushing them. Alerts in your account suggesting you set one. New criteria added. Bid multipliers, the old API-level way of doing this, are being retired, and advertisers who want to keep expressing value this way are required to move to value rules by 2027.
The reason is an information gap, and once you see it the contradiction resolves. Meta's delivery system is better than you at almost everything, but there are things it structurally cannot know. It does not know your margins. It does not know that the lead who asked about commercial work is worth eight times the one who asked about a repair. It does not know which of last year's customers are still with you. You know all of it, and value rules are the channel Meta built for you to say it in a language the auction understands.
That is also why the feature keeps growing. Meta is testing labels on custom audiences, so you can mark a list as qualified or disqualified leads, as high or low value customers, as recent purchasers or as people showing signs of churning, and then bid up or down on those labels. Feed your CRM's judgment back into the auction. For a business whose lead quality varies wildly, that is a more interesting prospect than anything in the targeting section ever was.
What to Do This Week
Take one campaign that has produced real business and break the last ninety days down by age, gender, placement and location. Look for one segment with real spend behind it and a real difference in outcome.
If you find nothing, stop. Nothing here will help you yet, and your money is better spent on the conversion data you are sending Meta in the first place, which is a deeper and more common problem than the bidding.
If you find something, set one rule, adjust it by what your own numbers justify rather than by 50 percent because it was offered, and put your most specific rule at the top of the list. Then leave it alone long enough to learn something, and judge it on the jobs you booked rather than the cost per lead you paid. That last part is the discipline the whole thing depends on, and it is the same discipline that separates an account chasing cheap form fills from one producing customers.
The instinct to reach for the cheapest possible result runs deep, and Meta will satisfy it every time you express it. It is worth remembering that an aggressive cap does not protect a budget either. The account that wins is not the one paying the least per lead. It is the one that has managed to explain, in terms the auction can act on, which customers were worth having.
Most accounts never get around to explaining it. If you would rather have someone do that work properly inside your account, our Facebook ads team does this for a living, and the first thing we would look at is the same breakdown described above.
