$2,000 a month is a real budget. It is also small enough that one bad split wastes most of it. Here is how I would divide it, what each slice actually buys at that number, and the point below which paid media stops being able to learn anything at all.
One thing this article is not going to do is re-argue which channel to fund first. That question has its own answer, in content marketing versus paid ads. This picks up after the decision is made and you are staring at a number that has to cover everything.
The split
All figures in Canadian dollars, per month.
Foundation, $400. The unglamorous slice. Page speed fixes, a Google Business Profile that is genuinely complete, title tags and meta descriptions on your five most important pages, and tracking that actually records what you think it records. This is the slice people cut first and regret first, because every other dollar you spend lands on the pages this slice repairs.
Content, $500. Realistically one solid page a month, or two shorter ones. Not a publishing schedule, a page that answers a question a buyer actually types. At this budget you are building an asset base slowly, so choose pages with buying intent over pages with traffic.
Paid media, $800. Roughly $26 a day. Read the next section before you assume that is enough, because it might not be, and the reason is arithmetic rather than opinion.
Tooling and measurement, $300. Email platform, scheduling, call or rank tracking if you genuinely read it. Go through the list once a quarter and cancel anything nobody has opened in sixty days. Dormant subscriptions are the quietest leak in a small budget.
The honest floor for paid media
Automated bidding runs on conversion data. Not much of it, but not none. Google's own guidance for Target CPA bidding is that "For evaluation, we recommend you measure performance for the last 30 days, including at least 30 conversions" (sourced from Google Ads Help).
Now do that arithmetic against your own numbers rather than an industry average, because your cost per lead is the only one that matters here. If a lead costs you $50, $800 a month buys about 16 conversions in 30 days, which is short of the 30 Google recommends for evaluation. If a lead costs you $120, you are at about 7, and the system is guessing.
There are two ways out of that, and only one of them is spending more.
Narrow the campaign until the money concentrates. One service, one geography, your best-converting keywords. A small budget spread across everything learns nothing about anything. A small budget pointed at one thing can actually optimise.
Count an earlier action as the conversion. A form start, a call over 60 seconds, a booking page view. You get more data points to work with, and you accept in exchange that each one is a rougher signal of real intent.
If neither of those gets you over the line, paid media at $800 is a test rather than a channel. That is a perfectly reasonable thing to run. Just budget it as a test and judge it as one, on a fixed window and a decision at the end, rather than letting it drift on for a year producing numbers nobody can read.
What to keep in-house at this budget
At $2,000 you cannot outsource everything, so outsource the work where an hour of expertise genuinely beats an hour of effort.
Outsource: technical fixes, campaign structure, conversion tracking setup, and the one content page.
Keep in-house: photos, review replies, profile posts, and answering the phone quickly. Nobody outside your business does these better than you do, and paying somebody else to do them badly is the most common way a small budget disappears.
This answer changes at around $5,000 a month. At that number, outsourcing the recurring work starts to pay, because what you are buying shifts from a one-off fix to consistency you would not otherwise sustain. At $2,000 you are buying expertise for the things you cannot do yourself, and doing the rest. If you want a real number rather than a guess for the outsourced half, our plans are priced publicly.
The checkpoints that mean re-allocate, not add
Every 90 days, three questions. They take twenty minutes and they are the difference between a budget and a standing order.
Is the foundation slice still finding things to fix? If it is, keep funding it. If the site is genuinely healthy, move that $400 into content or paid rather than inventing work for it.
Is any content page producing enquiries? If one is, write more like that one. If none are after two full quarters, the problem is topic choice, not volume, and writing twice as many of the same page will not fix it.
Is paid media above the conversion floor? If it is, and the cost per lead is stable, adding budget is a reasonable next move. If it is not, adding budget without narrowing the campaign just buys more noise.
The common failure here is not a bad split. It is a split set once in January and never questioned again. $2,000 a month allocated deliberately and reviewed quarterly will beat $3,000 spread evenly and forgotten, every time.
So which of your four slices has not been questioned since you set it?