Podcast: Download (Duration: 54:52 — 75.6MB)
Leave a new Facebook ad alone for at least 72 hours before judging it, and aim for roughly 1,000 reach before drawing conclusions. Killing an ad after a few hours is the most common mistake Rick Mulready sees, because the algorithm needs time to find your audience.
Rick has worked in online advertising for about 17 years, starting at AOL, then Yahoo, before going independent in 2012. He runs The Art of Paid Traffic podcast and teaches Facebook advertising to course creators.
His benchmarks for a webinar funnel are specific. Under $8 per registration is respectable for an online course, click through rate should clear 1%, and an opt-in page should convert at 20% or better.
This episode covers what to define before opening Ads Manager, how much to spend starting out, the troubleshooting metrics that explain a bad campaign, and how to structure ad sets so Facebook does not pick a winner too early.
A note on timing: this is a 2017 conversation. Facebook’s interface, objectives, and relevance scoring have changed since, so treat platform specifics as historical and the diagnostic approach as durable.
Get My Free Mini Course On How To Start A Successful Ecommerce Store
If you are interested in starting an ecommerce business, I put together a comprehensive package of resources that will help you launch your own online store from complete scratch. Be sure to grab it before you leave!
Table of Contents
Key takeaways
- Wait at least 72 hours and about 1,000 reach before judging any new ad.
- Under $8 per webinar registration is a reasonable target for most online course businesses.
- Calculate earnings per lead by dividing revenue by leads, which tells you what you can afford to pay.
- Aim for 1% or higher click through rate and 20% or higher opt-in page conversion.
- Relevance scores of 7 to 10 are strong, 4 to 6 warrant watching, and 1 to 3 need changes.
- Combine similar interests into one ad set to reach 500,000 to 2 million people and feed the algorithm.
- Run one ad per ad set, since Facebook declares a winner within hours and starves the others.
- Keep frequency between 1 and 4 to avoid banner blindness.
What should you decide before running Facebook ads?
Define your goal, your target audience, and your strategy before opening Ads Manager. Rick’s first question is always why, and hearing that friends are getting results is not an answer.
Audience definition goes well past the obvious. For a yoga product, women aged 25 to 45 interested in yoga is where most people stop.
The useful layer is adjacent behavior. What do they read, where do they shop, what do they watch, what conferences do they attend, what associations do they belong to.
Concretely that means Yoga Journal, Lululemon, Lorna Jane, Whole Foods. Those interests reach the same person through a different door.
Expect the picture to keep evolving. Rick has done this for years and still refines his own audience definitions, which differ product to product.
This preparation is the work nobody wants to do. It is also what separates campaigns that work from campaigns that burn money.
How much should you spend on Facebook ads starting out?
Whatever you can afford, and $5 to $10 a day is enough to generate usable data. Start small and reinvest as revenue appears.
The progression is gradual. Ten dollars a day is $300 a month, and returns get fed back until you are at $500, then $700, then $1,000.
Eventually the number stops being a guess. Once you know a dollar in returns two or three dollars, budget becomes arithmetic.
That requires history. You need campaigns behind you before you can spend confidently.
How do you calculate what you can pay per lead?
Divide total revenue by total leads to get earnings per lead. That figure is your ceiling on cost per registration.
The calculation is straightforward once a campaign has run. Revenue from a launch divided by registrations gives a number like $12.50.
Then the next campaign has a real budget. You can spend up to that figure per lead before losing money.
Rick’s general benchmark is under $8 per webinar registration for most course businesses. Lead costs vary by niche and by offer, and a free PDF download will always be cheaper than a webinar signup.
How long should you run a Facebook ad before judging it?
At least 72 hours, and roughly 1,000 reach where budget allows. Checking at ten in the morning after a six o’clock start tells you nothing.
Confirm it is running, then leave it alone. The algorithm needs time to work, and interfering resets that process.
Look at your objective metric first. For a webinar funnel, that is cost per registration.
Patience is what separates people who succeed at paid traffic. Those who keep testing past the point where they want to quit are usually the ones for whom it starts working.
What metrics diagnose a failing Facebook campaign?
Landing page conversion rate, click through rate, relevance score, and frequency. Rick calls these troubleshooting numbers, used after your primary objective metric shows a problem.
Click through rate should clear 1%. Anything around 0.85% is worth investigating against the other metrics.
Opt-in page conversion should hit 20% or better. This is registrations rather than sales.
Frequency measures how often one person sees your ad. Between 1 and 4 is healthy, and higher numbers produce banner blindness.
Smaller audiences drive frequency up faster. That is the mechanical reason narrow targeting fatigues quickly.
What is a good Facebook relevance score?
Seven to ten is strong, four to six is middling, and one to three usually needs changes. The scale runs one to ten and reflects how relevant Facebook judges your ad to be for your chosen audience.
Higher scores lower your costs and increase delivery. Facebook rewards ads it considers good content.
The score is not decisive on its own. Rick has seen ads with a relevance score of three producing perfectly good cost per lead.
When that happens, let it run. A campaign hitting your cost target is working regardless of the score.
Diagnosing a low score means examining both halves. The problem sits in either the ad itself or the audience you paired it with.
How should you structure Facebook ad sets?
Combine similar interests into one ad set to reach a potential audience of 500,000 to 2 million. Larger audiences give the algorithm more data to optimize against.
This reverses older advice. One interest per ad set gave cleaner reporting and exhausted a 60,000-person audience quickly.
Group interests that genuinely belong together. Wall Street Journal and New York Times share an ad set, and Whole Foods and Yoga Journal do not.
Warm audiences follow the same logic. Email lists and website visitors get combined to build sufficient size.
The mechanism is the pixel learning. Facebook studies who converts and finds more people resembling them, which requires volume.
Why you should run one ad per ad set
Facebook declares a winner within hours and starves the other ads before they have a fair test. Rick has tested this for about two years.
Three ads in one ad set produces a fast verdict. Within a few hours the algorithm picks one and gives it nearly all delivery.
Some of those losers were genuinely good. They never accumulated enough data to prove it.
The workaround is duplicate ad sets. Identical targeting, one ad each, which distributes delivery more evenly.
It is not a pure split test. It is close enough to see which creative actually performs.
Should you use images or video in Facebook ads?
Test both, and video carries advantages worth understanding. Facebook prioritizes video in the feed, so it gets more reach.
Connection is the first benefit. People see and hear you in a way a static image cannot deliver.
Engagement audiences are the strategic benefit. You can build audiences from people who watched 75% of a video versus 25%, then create lookalikes from them.
Production value matters less than relevance. A smartphone photo or a Facebook Live recording works when it speaks to the right audience.
Match the creative to the segment. A beginner and an advanced practitioner need different conversations, as do people reached through a magazine interest versus a retailer interest.
Which Facebook campaign objective should you choose?
Pick the objective that matches what you actually want to happen. Facebook optimizes toward whatever you select.
Conversions fits a direct webinar registration. You are telling Facebook to find people likely to complete the signup.
Traffic fits sending people to free content. You want clicks to the article, and any opt-ins are a bonus.
Traffic is meaningfully cheaper for content pages. The page is a reading experience rather than an opt-in experience, so optimizing for conversions fights the page’s actual purpose.
Video views fits awareness plays. You want people watching, and the retargeting audiences you build from watch time are frequently the real objective.
How do you warm up cold traffic before selling?
Send cold audiences to free content first and retarget them toward your offer. Running ads directly to a sale is the more expensive path.
Facebook is an interruption rather than a search. Nobody is looking for your product, so your ad has to earn attention and be relevant.
Rick’s analogy is dating. You do not propose to someone you just met.
Free content builds the relationship. A blog post or video positions you as an expert and gives people something they can use immediately.
Retargeting catches the ones who did not convert. People who consumed the content without opting in are a warm audience you can bring back to the registration page.
The funnel then extends further. Retargeting, engagement audiences, and Messenger conversations move people who are undecided toward a purchase.
What should a webinar landing page look like?
Match it to the ad that sent people there. Inconsistency between ad and landing page is a major cause of poor conversion.
Carry the same image across. If your ad used a photo, put that photo on the page.
Headlines and copy should echo too. The visitor needs immediate confirmation they are in the right place.
Color scheme counts as well. Anything that breaks continuity introduces doubt.
Use a purpose-built tool rather than building from scratch. Lead Pages, ClickFunnels, and Instapage all supply templates informed by their own testing.
Frequently asked questions
How much should a webinar registration cost on Facebook?
Under $8 is a reasonable target for most online course businesses. The right figure depends on your niche and offer, and a free PDF download will always cost less than a webinar signup.
How long should you run a Facebook ad before deciding?
At least 72 hours, and roughly 1,000 reach where your budget allows. Judging within hours is the most common and most expensive mistake.
What is a good click through rate on Facebook ads?
One percent or higher. Around 0.85% is worth diagnosing against your other metrics before making changes.
What is a good relevance score?
Seven to ten is strong and one to three usually needs changes. An ad hitting your cost per lead target is working even with a low score.
How much should you spend on Facebook ads starting out?
Five to ten dollars a day generates usable data. Reinvest returns gradually rather than committing a large budget before you have any history.
Should you put multiple ads in one ad set?
No. Facebook declares a winner within hours and starves the rest, so duplicate the ad set and run one ad in each for a fairer comparison.
How many people should a Facebook ad set target?
Between 500,000 and 2 million by combining similar interests. Larger audiences give the algorithm more data to optimize against.
What ad frequency is too high?
Above four. Frequency measures average views per person, and repeated exposure produces banner blindness, especially with smaller audiences.
