Digital Marketing Agencies

Ecommerce Advertising Agency: What to Actually Expect Before You Sign Anything

15+ years in local marketing; Google Ads certified; Shopify Partner.

You're running an online store. You're spending money on ads. And somewhere between the ad click and the checkout, something is leaking, but nobody can tell you exactly where.

That's the conversation I have with ecommerce owners pretty regularly. They've hired an ecommerce advertising agency, sometimes two, and they're still not sure if the ads are actually making them money or just keeping them busy. The reports look fine. The agency sounds confident. But the numbers on the P&L don't move the way they should.

This article is about what a real ecommerce advertising engagement looks like, what it costs, what work actually happens, and how to tell if you're getting it. I'm not going to cover every possible channel or platform in detail here. For a broader look at the agency landscape across Canada, our guide to digital marketing companies in Vancouver covers the market context well. What I want to do here is get specific about ecommerce advertising specifically, because it's its own thing and it deserves its own honest treatment.

What "Ecommerce Advertising" Actually Means (And What It Doesn't)

Ecommerce advertising is paid media, specifically built around driving product purchases online. That means Google Shopping campaigns, Performance Max, Meta product catalogue ads, and sometimes TikTok or Pinterest depending on what you sell.

It's different from general brand advertising. You're not trying to build awareness. You're trying to get someone to add something to a cart and check out. Every dollar spent has a clear job.

Here's the thing though: a lot of agencies pitch "ecommerce advertising" but they're really doing general paid search with a Shopify client. That's not the same thing. Real ecommerce advertising means your product feed is clean and structured, your shopping campaigns are segmented by margin and intent, your remarketing audiences are built from actual purchase behaviour, and your attribution model accounts for the fact that someone might see an ad three times before they buy.

If an agency can't tell you what your ROAS (return on ad spend, basically how many dollars you get back for every dollar you put in) was last month by product category, that's a signal worth paying attention to.

What This Should Actually Cost in Canada

Canadian ecommerce businesses typically pay somewhere between CA$1,500 and CA$5,000 per month in agency management fees, with ad spend on top of that. Per the pricing benchmarks available from Canadian agency directories in 2024, mid-market retainers for paid media management run CA$3,000 to CA$10,000 monthly, depending on complexity and account size.

Here's a worked example. Say you're spending CA$5,000 per month on Google Shopping and Meta ads combined. A reasonable management fee is somewhere between 15-20% of ad spend at the lower end, or a flat retainer in the CA$1,500 to CA$2,500 range. So your total monthly investment is roughly CA$6,500 to CA$7,500. At a 3x ROAS, that CA$5,000 in ad spend should be generating CA$15,000 in revenue. Your agency fee needs to make sense relative to that number. If you're paying CA$3,000/month in fees to manage CA$5,000 in spend, the math gets uncomfortable fast.

One thing worth knowing: Canadian Google Ads CPCs are generally 30-50% lower than US equivalents for most ecommerce categories (per DataForSEO Canadian keyword data, 2024). That means your budget goes further here than it would for an American competitor running the same campaigns. A good agency should be pointing that out to you, not hiding it.

What the First 60 Days Should Actually Look Like

This is where I see the most confusion. Ecommerce advertising isn't a flip-a-switch situation. There's real work that has to happen before the ads perform well, and if an agency skips it, you'll pay for it later in wasted spend.

Weeks 1-2: Feed and foundation audit. Before anyone touches a campaign, your product feed needs to be reviewed. That means checking titles, descriptions, GTINs, pricing accuracy, and whether your product categories map correctly to Google's taxonomy. A messy feed is the single most common reason Google Shopping underperforms. In my experience, most ecommerce accounts I've audited have at least 20-30% of their feed either miscategorised or missing attributes that affect ad eligibility.

Weeks 3-4: Pixel and tracking verification. This is not glamorous work. It's checking that your purchase event fires correctly, that your cart abandonment audiences are building, that GA4 is recording transactions accurately, and that your Meta pixel isn't double-counting conversions. I've seen accounts where the reported ROAS was 4x but actual revenue from ads was closer to 2x because of tracking errors. Fixing this before you scale spend is critical.

Month 2, Weeks 1-2: Campaign build or restructure. Now you actually build the campaigns. Shopping campaigns should be segmented by margin tier if possible (high-margin products deserve more budget), with separate ad groups for branded vs. non-branded queries. Performance Max campaigns need audience signals fed in from your actual customer list, not just Google's guesses. Meta campaigns need a cold-audience creative test running alongside a warm-audience retargeting campaign.

Month 2, Weeks 3-4: First real data review. You look at what's working, what's wasting money, and where the conversion rate is dropping off. This is also when you have the first honest conversation about whether the product margins support the ad costs at all. Sometimes they don't, and that's information you need.

Typically, accounts that go through this process properly start to show consistent ROAS data by the end of month two. Accounts that skip the feed and tracking work often spend three to four months optimising campaigns that were broken from the start.

The Attribution Problem Nobody Talks About Upfront

Here's the piece that most agency pitches gloss over: ecommerce attribution is genuinely hard, and any agency that doesn't acknowledge that upfront is either naive or not being straight with you.

The problem is that a customer might click a Google Shopping ad, leave, come back through a direct visit two days later, and buy. Google Ads takes credit. Meta takes credit if they also saw a Facebook ad. Your Shopify dashboard shows a direct sale. Three different numbers, one transaction.

A good ecommerce advertising agency will set up a consistent attribution model and stick to it, then explain the logic to you clearly. Usually that means last-click for Google Shopping (conservative, easy to defend), and view-through attribution turned off or set to one day for Meta (because the default seven-day view-through makes Meta look much better than it is).

In my experience, when you tighten attribution to a defensible model, reported ROAS often drops 20-40% from what the agency was showing before. That's not a failure. That's honesty. The ones who resist this conversation are the ones to worry about.

If you're using email platforms alongside paid media, the attribution question extends there too. For how email tools like Klaviyo fit into the picture, our Klaviyo agency overview covers what that layer looks like in practice. And if you're on Mailchimp, the same questions apply, see our Mailchimp agency breakdown for what to expect there.

Red Flags to Watch Before You Sign

They quote ROAS without telling you how it's measured. ROAS means nothing without the attribution window and model attached to it. "We deliver 4x ROAS" is a claim that can mean almost anything.

They own your ad accounts. Your Google Ads account, your Meta Business Manager, your Google Merchant Centre. These should all be yours, with the agency added as a user. If an agency creates the account under their own umbrella, you can't take your history with you when you leave. Per the Competition Bureau's guidance on misleading business practices (Competition Act, sections 52 and 74.01), agencies that obscure account ownership or make it difficult to leave may have other transparency problems too.

The pitch has no numbers from past clients. Not even anonymised ones. "We've helped ecommerce brands grow" is not a case study. Ask for cost per acquisition, ROAS, and revenue attributed over a defined period. If they won't show you the numbers, they probably don't have good ones.

They're not asking about your margins. An agency that doesn't know your product margins can't tell you whether your ad spend is actually profitable. A 3x ROAS looks great until you find out your margins are 25% and you need a 4x just to break even.

Month-to-month is resisted. Any agency worth working with should be willing to operate without a long-term lock-in. The work should speak for itself.