We have already covered how much to put into marketing overall. But once you have your number — say $3,000 a month — the real question starts: where does it go? All into Meta ads? Some SEO? Content? A newsletter? This is where most SMBs get it wrong, and it costs them for months without their noticing.
How you split it matters more than the amount. A modest budget well allocated beats a large one badly allocated every time. Here is how we carve it up when advising a Quebec SMB, and why.
Three families of channels, three jobs
Before talking percentages, understand that your channels are not doing the same job. Paid acquisition (Google Ads, Meta, LinkedIn) buys attention now: you pay, it comes, you stop paying, it stops. That is renting.
SEO and content (articles, service pages, your Google profile) build an asset: it takes months to come up, but once it is in place it brings customers without you paying for the click. Email retention works your existing base: people who already know you, vastly cheaper to win over again than a stranger. The classic mistake is putting everything into the first family because it is the only one that shows a result the following week.
The 70-20-10 rule
A proven split for deciding without going in circles: 70% on what already works and that you measure, 20% on what is promising but not yet proven in your business, 10% on experiments you accept you may lose. Concretely, if Google Ads reliably brings you contracts, that is your 70%. The 20% might be starting to publish substantial articles and tightening your SEO — a serious medium-term bet. The 10% is testing a new network, a video format, a retargeting campaign you have never tried.
70-20-10 is not a magic formula, it is a discipline. It forces you to protect what works while keeping a door open on your next main channel.
Adjust for your sales cycle and your maturity
70-20-10 is a starting point. Two things shift it. First, your sales cycle. An industrial B2B closing over six months is better off leaning towards SEO, content and email: its prospects research at length, compare, come back. A renovation service with a fast decision benefits more from paid acquisition, because the demand is immediate and local.
Second, your maturity. An SMB starting from nothing, with no traffic and no list, first has to buy visibility in order to exist: early on, advertising can legitimately take 70 to 80%. As your content climbs in Google and your list grows, you shift budget towards the assets you own rather than rent. It is a gradual slide, not a single dramatic switch: you reduce advertising at the pace the other channels take over.
The two mistakes that sink an allocation
The first is putting everything into advertising. As long as you pay, leads come in, and the illusion of control is pleasant. The day you cut the budget, or the cost per click climbs, it all stops dead. You have built nothing: you rented for two years without ever buying.
The second is measuring nothing. If you do not know which channel brings actual customers (not clicks, customers), you are not allocating, you are guessing. Before talking about 70-20-10, you need at minimum to know where your last ten customers came from. Without that, any allocation is a roll of the dice.
An example: a services SMB at $3,000 a month
Take an established B2B services firm with some traffic and a small list. A realistic split: $1,800 into Google Ads and retargeting (the 70%, what brings in mandates today), $600 into content and SEO, meaning two substantial articles a month plus optimizing the service pages (the 20%, the asset on the rise), and $300 into email and automation to re-engage older prospects (the 10%, the cheapest and the most profitable over time).
If you are unsure about your split, or you suspect your budget leans too far one way, that is exactly what we untangle during the free 15-minute diagnostic call. We look at where your money goes, what each channel really returns, and where a simple reallocation would win you customers without spending another dollar.
The one-page marketing plan template — where this allocation gets written down.