How Much It Costs to Advertise a Small Business

$300 a month. $6,500 a month.

Both numbers get called the average answer to how much it costs to advertise a small business, in guides published the same month.

Neither one is lying. Each is describing a different business, at a different stage, and rounding it down to a single headline.

Here's the real number, where it comes from, and what to do with it this week.

What It Actually Costs to Advertise a Small Business, Starting With Revenue

Start with revenue, not the channel.

The U.S. Small Business Administration has told small businesses for years to plan on 7% to 8% of gross revenue for marketing and advertising, for anyone under $5 million a year. That guidance is old. The CMO Survey, run every spring by Duke's Fuqua School of Business with Deloitte and the American Marketing Association, put the real 2026 average higher, at 9.0% of revenue across companies of every size, in its March 2026 topline report.

The two numbers disagree by about two points. Trust the higher one if your business is under three years old and still building the audience that later spending gets to reach for less. Trust the lower one once repeat customers are already doing some of the advertising's job.

A business doing $500,000 a year lands, at 7% to 9%, somewhere between $35,000 and $45,000 a year. That's $2,900 to $3,750 a month.

Every channel-specific figure in this post has to fit inside that number, not add on top of it.

Run the math for your own revenue right now. One multiplication. It's the only "how much should I spend" answer that starts from your business instead of somebody else's average.

That percentage covers media spend: buying the clicks and the airtime. It doesn't cover building what runs in that space. Gidds Media's 2026 breakdown prices a template-based ad or landing page at $500 to $2,000, a freelancer's custom version at $2,500 to $7,500, and a full agency production at $7,500 to $25,000. Budget for creative separately from the number above, or the media line quietly pays for both and neither one is fully funded.

What Each Channel Actually Charges in 2026

The percentage sets the size of the envelope. This table is what fits inside it, and more of this kind of number lives across the advertising shelf of this blog, priced by channel.

Two of those rows deserve a second look before you write a check.

Google's $5.42 average cost per click buys a search from someone already looking for what you sell, which is why the average cost per lead sits at $66.69 even though the click itself looks cheap on its own.

Meta's click is cheaper, but the two benchmarks disagree on how cheap. The eDigital tracks its own accounts at $0.78, up from $0.70 in 2025, while Stackmatix's broader 2026 sample averages $1.14. Trust The eDigital's figure for a like-for-like read on rising cost. Trust Stackmatix's if your account blends several campaign objectives, since a mixed sample pulls the average up.

If your number lands closer to the bottom of that range than the top, our cheap advertising for small business guide covers channels that undercut every row here.

Why the Average Hides the Number That Matters to You

An "average" cost per click blends an attorney paying $9.87 with an arts venue paying $1.63. Neither figure describes your business. Both feed the average quoted at you.

LocaliQ's 2026 industry breakdown shows the real spread:

If your industry isn't near the $5.42 blended average, budget against your own row, not the headline figure. Search "[your industry] average cost per click 2026" and cross-check two sources before you commit a number. It's a five-minute habit that changes the entire monthly figure you land on.

Cost per click also only measures what happens after somebody notices you. Cost per thousand impressions, CPM, measures what it costs just to be in front of them at all, whether they click or not, and the two rankings don't match. Top Draw's 2026 comparison puts Google Display at $3 CPM, Facebook at $8.60, a roadside billboard at $13 to $22, and a print magazine spread at $140 to $1,300. A magazine ad can cost forty times more to show than it does to get clicked on Google, because nobody's clicking a magazine.

That's the case for pairing a CPM channel with a CPC channel instead of picking one. The CPM channel buys the shape of the brand a stranger recognizes later. The CPC channel buys the click from someone who already recognizes it. $200 to $400 a month is enough to stay visible on a CPM channel in most local markets, small enough that it doesn't crowd out the click-based test that's actually measurable.

There's a calendar wrinkle too. Auction CPMs on Meta spike sharply in the fourth quarter as holiday advertisers pile into the same inventory, sometimes pushing the same click 30% to 50% higher than a summer month. If your annual budget is fixed, weight more of it toward fixed-price placements in November and December, and let the auction channels rest until January, when the same dollar buys more.

Setting Your Own Number: A Worked Example

Take a home services business doing $40,000 a month in revenue. At 8% of revenue, the SBA's baseline, that's $3,200 a month for advertising, total.

Split it like this:

1. $1,500 to Google Search Ads. Home services runs closer to $7 to $8 a click, above the Business Services row in the table, funding 190 to 215 clicks a month, enough to see whether the campaign converts anywhere near the 8.18% average conversion rate WordStream reports.
2. $700 to Meta retargeting, aimed only at people who already visited the site. At $0.78 a click that's roughly 900 clicks, plenty to warm up a list that size.
3. $500 to one placement with a fixed price instead of an hourly auction, a newsletter sponsorship or a leaderboard entry.
4. $500 held in reserve, spent only on whichever channel above shows a cost per lead under $70 by day 14.

Now take a smaller case. A solo business doing $8,000 a month lands, at 8%, on $640 a month total. That's below the $500 to $800 floor a single Google Ads test needs to mean anything, so splitting it across two channels tests neither one properly.

The fix isn't to spend less carefully. It's to spend the whole $640 on the one channel with the lowest cost per click for the business's category, and run it for a full month instead of two weeks.

At the Business Services row above, $5.87 a click, $640 buys about 109 clicks. At an 8.18% conversion rate, that's roughly nine conversions, enough for a real cost-per-lead figure instead of a guess. The second channel becomes next month's decision, funded by what the first one just proved.

Track the actual numbers as you go, not just the plan. Three columns are enough: spend, clicks, conversions, updated once a week. That habit turns next month's decision into arithmetic instead of a guess about which channel felt like it was working.

The Minimum Spend Before Any of This Data Means Anything

A $200 test on Google Ads at $5.42 a click buys about 37 clicks. At an 8.18% conversion rate, that's three conversions on a good month. Three data points tell you almost nothing about whether a channel works.

The floor for a Google Ads test that actually means something is closer to $500 to $800: 90 to 150 clicks and enough conversions to compare honestly against the $66.69 average cost per lead. Below that, a failed test and an unlucky week look identical. Our cost-per-visitor breakdown found the same floor holds across paid traffic generally, not just search.

Meta's cheaper click changes the math in your favor. $150 at $0.78 a click buys roughly 190 clicks, enough for a real read on click-through rate before anything even converts.

A streaming TV test, run for two weeks at Adwave's roughly $50-per-spot floor, costs $700 to $1,000 depending on frequency. That's why it's usually the third channel added, not the first: it needs a media budget already proven elsewhere before it earns a slot.

Spend less than the floor on any of these and you're not testing the channel. You're buying a rounding error and calling it a result.

What to Cut First When the Budget Doesn't Stretch

Every budget eventually gets squeezed, and the instinct is to cut evenly across every channel. That's the wrong move.

Cut the channel with the fewest data points first, not the most expensive one. A channel you've run for two weeks has no track record worth protecting. A channel that's already produced ten leads at a known cost has earned another month.

Set the kill signal before you need it. Past 100 clicks with no measurable cost per lead, or past 200 clicks at double your category's average CPC, that channel goes first, whatever its age.

Write that number down somewhere visible before you need it, not just in your head. A rule you have to remember under pressure is a rule you'll bend the day the numbers look almost good enough.

Fixed-price placements belong in the protect column longer than auctions, because their price doesn't move week to week. Our paid placement for small brands breakdown covers five of these, priced against each other.

The Board runs the plainest version of a fixed-price row. Money is the only ranking signal it uses, stated next to every listing rather than hidden in a media kit, and right now taking the top spot costs $103, a number that only rises until somebody wins the current season. The advertising for $1 entry price is real too, and it's the cheapest row in the whole table above.

The Number to Write Down Today

Multiply your monthly revenue by 0.07, then by 0.09. That range, not the $300 or the $6,500 you saw first, is your actual budget.

Split it across two channels with a track record and one placement with a fixed price. Hold back 15% as reserve. Check cost per lead on day 14, not day 30.

The number changes every year. The method for finding it doesn't. Run it again next quarter, because the guides quoting a fixed monthly figure as gospel are the ones worth trusting least.