A restaurant owner spends $400 on Facebook ads over two weeks, gets forty clicks to her website, three phone calls, and no reservations. She turns the ads off and tells the next owner at the chamber of commerce meeting that Facebook ads don't work for restaurants. What actually happened is that nobody tested whether her website could turn a visitor into a reservation in the first place, and two weeks was never enough time to know anything.
This is the mistake, and it happens before the ad, the budget, or the platform ever get chosen. An SBA-affiliated small business seminar describes its own pitch this way: "Before you spend a dollar on Facebook or Google ads, spend one hour learning what nobody selling ads will tell you." The seminar's framing treats whether to advertise at all as a separate decision from which platform to pick, and that separation is the part most owners skip.
Your website has to prove itself first.
Before any ad runs, your site or landing page is already getting some traffic, from search, social, word of mouth, or people typing your name into Google after a friend mentioned you. That traffic is a free readiness test, and most owners never look at what it's telling them.
An independent marketing consultant with two decades running ad accounts puts a number on what "working" looks like, though the figure is one experienced practitioner's benchmark rather than a universal industry standard: "A conversion rate of 1-3% is realistic for a well-set-up landing page." That means out of every 100 people who land on your page, one to three should book, buy, or call. If your organic traffic isn't hitting that now, sending more visitors of the same kind won't fix the underlying problem — you'd just be paying to send more people to the same leaky page.
Separate research on landing pages backs up why that number matters and where it usually breaks. Sites with a single clear next step, one button, one offer, convert far better than sites offering visitors three or four choices at once. A roundup of landing page data found that "landing pages with a single CTA—call to action, the button or line that tells someone what to do next: book now, call today, get a quote—convert 266% better than pages with multiple competing calls to action." If your homepage is trying to sell three things and explain your whole history at the same time, that's likely your problem, and it costs nothing to fix before you ever touch an ad account.
One uncomfortable finding from the same review: mobile pages convert at roughly 4.1%, compared to 6.3% on desktop, despite mobile carrying 60% of the traffic. If most of your customers browse on their phone, and your site is slow or hard to tap through on a phone, that gap is quietly eating your ad spend before a single ad is judged good or bad. Fixing your mobile experience matters, but even a fast site can't tell you whether the ad itself is strong, and it's the problem we're building Knowing to solve.
How fast you respond matters as much as the ad itself.
Say the ad works. Someone clicks, likes what they see, and fills out a form or calls. What happens in the next hour decides whether that interest turns into money.
The invincible brands piece names this directly as one of three things to check before spending anything: a fast, consistent follow-up process, distinguishing a business that responds within an hour from one where a lead "sits 3 days." A three-day gap isn't a minor delay. It's usually the difference between a customer and a stranger who found somebody else in the meantime.
This matters because it's invisible in ad account dashboards. The platform will report your click-through rate and cost per click while the actual leak, the unanswered voicemail, the contact form nobody checks on weekends, sits entirely outside what the ad account can see.
A short test isn't a fair test.
The restaurant owner's mistake wasn't just the website. It was judging two weeks and forty clicks as a verdict.
The same practitioner source sets a floor: "Most campaigns need four to eight weeks of consistent spend before meaningful conclusions can be drawn." Ad platforms need time to figure out who responds to your ad and show it to more people like them; cutting the test short means you're judging the campaign during the part where the platform was still learning, not during the part where it actually performed.
Volume matters as much as time. The same source gives a rough floor for that too: "you need enough monthly budget to generate at least 50-100 clicks or meaningful impressions to your target audience before drawing conclusions." Forty clicks over two weeks clears neither bar. It's not that the ad failed; it's that there wasn't enough evidence yet to say anything.
The ad platform will always tell you something happened. It won't tell you whether what happened means anything.
The fine print that catches people off guard.
Even the introductory offers meant to lower the barrier to entry have terms that trip owners up. The SBA seminar's description says it will cover Google's new-advertiser credit, promising to explain "what 'spend $500, get $500' really requires, so you're never surprised by your card statement." The offer sounds like free money; it comes with spend minimums, time windows, and eligibility rules that most people don't read until the bill arrives. That's a small example of a bigger pattern: the businesses selling you ad space have no reason to walk you through when advertising isn't the right move yet.
Sometimes the honest answer is not yet.
If your landing page can't hit 1 to 3%, if follow-up takes days instead of minutes, and if your budget can't sustain six to eight weeks of consistent spend, running ads right now mostly converts your marketing budget into evidence that you weren't ready. That's a real cost, and it's the one nobody selling ad space mentions.
A piece from marketing consultancy Invincible Brands makes a case for the alternative worth taking seriously: for a locally targeted business without the budget to sustain a proper test window, consistent free content, posted regularly where your customers already look, can outperform a short paid burst that gets shut off before it had a chance to prove anything. That's not a consolation prize. It's often the more honest use of a limited marketing budget while you fix the landing page, tighten the follow-up, and build up the cash to run a real test later.
The checklist, in the end, isn't about Facebook versus Google, or which budget tier to pick. It's about whether the business behind the ad is built to catch what the ad sends it. Fix that first, and the ad decision gets a lot simpler, because you'll actually be able to tell whether it worked.
Frequently asked
Common questions.
How long should I run ads before deciding if they're working?
Give a new campaign four to eight weeks of consistent spend and aim for at least 50 to 100 clicks before judging results. Shorter windows usually just capture the period where the ad platform is still learning who to show your ad to, not real performance.
What conversion rate should my website have before I start advertising?
One experienced ad consultant suggests a well-built landing page typically converts 1 to 3% of visitors into leads or sales, though the right number varies by industry. Check your current organic traffic against that benchmark first; if your site isn't hitting it without ads, paid traffic will likely convert at a similar rate or worse, since the underlying page is the bottleneck.
Is it ever better to skip paid ads entirely?
Yes. If your budget can't sustain a proper four-to-eight-week test, or your website and follow-up process aren't ready to convert traffic, steady free content on channels your customers already use can outperform a short paid campaign you can't fund long enough to judge fairly.