📚 What You Will Learn Today:
When you set up a Facebook ad, there's a little question hiding in the settings that decides whether your money works hard or gets wasted: "What are you optimizing for?" Most business owners click whatever is already selected and move on. That one click can be the difference between cheap clicks from the wrong people and real paying customers. This guide explains it all in plain words.
By the end of this guide, you will know:
- What "optimizing" actually means (it's simpler than it sounds)
- The 5 things you can tell Facebook to go find, from clicks all the way to clients
- Why asking for the biggest fish isn't always the smartest move
- The three bidding choices and which one fits your business
- A shortcut if you don't have time to do it all
🤔 First, What Does "Optimize" Even Mean?
Here's the simple version: Facebook is a very smart delivery robot. When you set your ad to "optimize for" something, you're telling the robot what to go find.
Think of it like sending a dog to fetch. If you say "fetch a stick," the dog brings back any stick. If you say "fetch the red ball," it hunts for the red ball. Facebook works the same way. Tell it "find me people who click," and it finds clickers. Tell it "find me people who buy," and it looks for buyers.
The catch: Facebook can only learn what you want if you show it examples. The robot needs to see a bunch of "red balls" before it knows what one looks like. That's why the rarer the thing you ask for, the harder it is for Facebook to learn. We'll come back to that.
A dentist ran the same ad two ways. Version A was set to "link clicks." Version B was set to "leads." Version A got clicks for 40 cents each, and the dentist was thrilled. But only 2 people booked. Version B got clicks for $1.80 each, which looked worse, but 14 people booked. Same ad, same budget. The robot was fetching different things.
The Pixel Is the Robot's Eyes
To know who did what, Facebook needs a tiny piece of code on your website called the pixel (and its newer partner, the Conversions API). Every time someone fills out a form, books a visit, or buys, the pixel tells Facebook "this one did the thing!" Without it, the robot is fetching blind. Get it installed before you do anything else in this guide.
📊 The Ladder of Things You Can Ask For
You can point the robot at any rung. The right one depends on how many examples you can give it.
✅ The 5 Things You Can Optimize For
Let's climb the ladder one rung at a time, from the easiest to the most valuable.
Optimizing for Link Clicks
This tells Facebook: "Find me people who like to tap on ads." And it will. You'll get lots of cheap clicks and your report will look amazing.
The problem? Some people click on everything. They tap, glance, and leave. Facebook will happily find thousands of these people because that's exactly what you asked for. Cheap clicks that go nowhere.
When It Makes Sense
- You're brand new and have zero data. You need traffic just to get the pixel some examples.
- You're sharing a helpful blog post and don't need a form filled out.
- You're testing which picture or headline people respond to.
Running a "clicks" campaign for months and wondering why nobody books. Clicks are the bottom rung. Use them to get started, then climb.
Optimizing for Leads
Now you're telling the robot: "Find me people who will give me their name and phone number." This is where most local businesses should live, at least at first.
Clicks will cost more. But a much bigger share of those clicks will turn into leads, because Facebook is now hunting for form-fillers, not tappers.
A chiropractor spending $30 a day switched from clicks to leads. Her clicks went from 50 cents to $1.50. Scary! But her leads went from $38 each to $11 each. She got more than three times as many people to call for the same money.
Two Ways to Collect Leads
- Instant forms (on Facebook): The form pops up right inside Facebook, already filled with their name and phone. Very cheap, very easy. Leads can be a little less serious.
- Your own landing page: They leave Facebook and fill out a form on your site. Costs more per lead, but people who make the extra effort tend to be more serious.
Whichever you pick, text every new lead within 5 minutes, automatically. A cheap lead that you call tomorrow is a wasted lead. Speed turns "kind of interested" into "booked."
Optimizing for Qualified Leads
Not all leads are equal. A gym gets a lead from someone who lives 40 miles away and just wanted the free guide. A realtor gets a "seller lead" from someone who's "maybe moving in 3 years." Those aren't really leads yet.
A qualified lead is someone who answered a few extra questions the right way. "Do you live within 10 miles?" "Are you looking to start in the next 30 days?" "What's your budget range?"
How to Tell Facebook About Them
When someone passes your questions, fire a special event to the pixel (call it something like "QualifiedLead"). Then set your ad to optimize for that event. Now Facebook stops hunting for anyone who fills out a form and starts hunting for people who look like your good leads.
Fishing with a net catches everything, including boots and seaweed. Adding qualifying questions is like using a net with bigger holes. The tiny stuff slips through, and what's left is worth keeping.
Every question you add drops the number of leads. Three questions is plenty. Ten questions and nobody finishes the form.
Optimizing for People Who Booked
One more rung up. Now you tell Facebook: "Only count it when they actually schedule a visit, a call, or an estimate." This is very powerful, because a booking is one step away from money.
How It Works
- Your booking calendar sends a "Schedule" event to the pixel when someone picks a time.
- You set your ad to optimize for "Schedule."
- Facebook studies everyone who booked and goes looking for more people like them.
The Catch
Remember the fetch dog? It needs to see enough red balls to learn. Facebook likes to see about 50 of your chosen event every week to get smart. If you only get 8 bookings a week, the robot stays confused. Your results jump around, and costs can get weird.
Not getting 50 bookings a week? Stay one rung lower (leads or qualified leads) where you have more examples. You can always climb later as your numbers grow.
Optimizing for Paying Clients: The Big Fish
The top of the ladder. You tell Facebook: "I only care about people who pay." If you can pull this off, your ads get spookily good at finding buyers.
What You Need
- A way to tell Facebook when someone paid. For online checkout, the pixel does it. For in-person sales (most local businesses!), you upload your customer list or connect your payment system so Facebook can match up who paid.
- Enough sales. This is the hard part. A med spa doing 15 sales a week doesn't have enough examples for Facebook to learn fast.
- Patience. The learning period takes longer, and the first week or two can look bad before it gets good.
When you track sales, count the sale on the day the customer paid, not the day they became a lead or booked. If you send Facebook the wrong date, the robot learns the wrong lesson about which ads brought buyers.
A martial arts school with three locations was signing up about 60 new members a month. That's enough to try the big fish. They uploaded paid memberships back to Facebook every day and switched their ads to optimize for purchases. Over two months, their cost per new member dropped by about a third, even though their cost per lead went up. They were paying more for fewer leads, but far more of them turned into members.
📊 Cheaper Leads Are Not Always Better
Judge your ads by the bottom line, not the top line.
Bidding: How Much Should You Tell Facebook to Spend?
Once you've picked what to fetch, Facebook asks how hard to chase it. This is called your bid strategy. There are really only three choices that matter for a local business.
Option 1: Lowest Cost (the default)
You say: "Here's my daily budget. Spend all of it and get me as many as you can." Facebook does its best. This is the right choice for most people, most of the time. It's simple and it lets the robot learn fastest.
Option 2: Cost Cap
You say: "Get me as many as you can, but try to keep the average under $X each." Good when you know exactly what a lead is worth to you. The risk: if your cap is too low, Facebook stops spending and you get nothing. It's like telling a shopper "find me steak under $2 a pound." They'll come home empty-handed.
Option 3: Bid Cap
You say: "Never bid more than $X in any single auction." This is the strictest. It gives you the most control and the most ways to accidentally shut your ads off. Skip it unless you really know your numbers.
A home services company knew a booked estimate was worth about $90 to them. They set a cost cap of $45 per booked estimate. Facebook delivered steadily at around $40. When they got greedy and dropped the cap to $25, spending fell to almost zero within two days. They moved it back to $45 and it came right back.
Start with lowest cost. Run it for two or three weeks. Look at what your leads or bookings actually cost. Only then, if you want more control, try a cost cap set a little above that number.
Limited on Time? Try This Shortcut
Don't have the hours to set up qualifying questions, booking events, and sales uploads? Here is the simple version that works for most local businesses.
- Install the pixel (or have someone do it for you). Twenty minutes, one time.
- Optimize for leads. It's the sweet spot: valuable enough to matter, common enough for Facebook to learn fast.
- Use lowest cost bidding. Don't overthink it.
- Text every lead within 5 minutes, automatically. This makes a lead campaign perform like a booking campaign, because your follow-up does the climbing instead of the robot.
- Once you pass 50 bookings a week, climb one rung and optimize for bookings.
That's it. Two settings and one text system. You'll beat 90% of businesses who are fiddling with knobs they don't understand.
📝 Quick Summary
Everything In One Place
- Optimizing = telling Facebook's robot what to go fetch
- The pixel is how the robot sees what people do. Install it first
- Clicks are cheap and mostly useless. Only for starting out or testing
- Leads are the sweet spot for most local businesses
- Qualified leads and bookings are better, but you need about 50 a week for Facebook to learn
- Paying clients is the big fish. Only try it when you have enough sales, and count each sale on the day they paid
- Lowest cost bidding for most people. Cost cap only once you know your numbers
- Fast follow-up makes any level work better
Remember: Ask the robot for what you actually want. If you ask for clicks, you get clickers. If you ask for customers, you get customers.
🎓 Your Action Plan
Week 1 — Give the robot eyes
- Install the Facebook pixel on your website and confirm it fires on your thank-you page
- Set up automatic texting so every new lead gets a message within 5 minutes
Week 2 — Pick your rung
- Launch one campaign optimized for leads with lowest cost bidding
- Leave it alone for 7 days. Don't touch it while the robot learns
Week 3 — Measure
- Write down: cost per lead, how many booked, how many paid (on the day they paid)
- Add 2 or 3 qualifying questions if too many leads are the wrong fit
Week 4 — Climb if you can
- Getting 50+ bookings a week? Switch to optimizing for bookings
- Know your cost per lead? Try a cost cap set slightly above it
- Not there yet? Stay on leads and keep improving your follow-up
The fanciest optimization setting in the world can't beat a plain lead campaign with a fast, friendly follow-up system behind it. Get the follow-up right first. Then let the robot help.
Ready to Get More Customers From Your Ads?
We set up and run Facebook and Google ads for local businesses every day. The ads, the tracking, the follow-up texts, and the booking are all done for you.
Get Your Free Growth Strategy Session