Short version: A good DTC Google Ads agency is not defined by its logo wall or its ROAS screenshots. It is defined by nine capabilities: bidding to profit, owning the product feed, splitting Performance Max by margin, keeping brand and non-brand apart, paying more for new customers on purpose, measurement you can trust, a steady creative supply, scaling in controlled steps, and reporting that you own. This piece walks through each one, the question that tests it, and the answer that should make you nervous.
Most DTC founders hire on vibes. The pitch deck looks sharp, the case studies have big numbers, the call goes well. Six months later the account is spending more, the ROAS in the dashboard looks fine, and the bank account does not agree. That gap almost always traces back to one of the capabilities below being missing.
This is not another ranked list. If you want names, our guide to choosing an ecommerce Google Ads agency covers ten options and a scorecard. This piece is the layer underneath: what the agency has to be able to do, and how to find out in one call whether it can.
Why DTC needs a different kind of Google Ads work
A DTC brand lives and dies on contribution margin. You are not a marketplace seller with someone else carrying the brand, and you are not a lead gen business where a form fill is the finish line. Every order has a product cost, shipping, payment fees, returns, and a discount code attached, and Google Ads has no idea about any of it unless someone tells it.
That is why so much generic PPC management quietly fails DTC brands. The account gets optimised to the number Google can see (conversion value) instead of the number the business runs on (profit after costs). The agency hits its target. The brand loses money on every incremental order. Everyone is technically right and the outcome is still wrong.
So when you evaluate an agency, ignore the slide about "data-driven strategy". Test the nine things below.
The 9 capabilities, and how to test each one
1. Bidding to profit, not just ROAS
ROAS treats a $100 order of a 70% margin product the same as a $100 order of a 20% margin product. A capable agency knows your break-even ROAS by product group, sets targets above it, and where the data allows, feeds margin or profit into the conversion value so Smart Bidding chases the right orders.
Ask: "What is our break-even ROAS, and how did you work it out?"
Worry if: they quote an industry average, or say "we aim for 4x" before they have seen your costs.
2. Owning the product feed
Shopping and Performance Max run on your product feed. Titles, product types, custom labels and GTINs decide which searches you show for and how cleanly the account can be split. The product feed is the most ignored part of Google Ads, and it is usually where the cheapest wins are hiding.
Ask: "Who edits the feed, how often, and what custom labels would you add for us?"
Worry if: the answer is "that is your developer's job" or "the Shopify app handles it".
3. Segmenting Performance Max by margin and price
One Performance Max campaign across the whole catalogue lets Google spend where it is easiest to get conversion value, which is usually your bestsellers and your brand terms. Splitting campaigns by margin band, price band or hero versus long tail puts budget where profit is. We have seen this single change carry an account: a USA pen brand grew 80%+ to $240K a month after price-segmented Performance Max, and a USA loungewear brand grew revenue 70% to $380K with ROAS at 9x.
Ask: "How would you split our catalogue across campaigns, and why?"
Worry if: "Performance Max works best with everything in one campaign so the algorithm has more data." Sometimes true for a tiny catalogue. Usually a sign nobody wants to do the work.
4. Keeping brand and non-brand apart
Brand searches convert at very high ROAS because those people were coming anyway. Blend them into prospecting campaigns and the headline ROAS looks great while real growth stalls. A capable agency reports brand and non-brand separately, protects brand cheaply, and judges growth on non-brand.
Ask: "What share of our revenue right now comes from people already searching our brand name?"
Worry if: they cannot answer it after an audit, or they have never looked.
5. Paying for new customers on purpose
Returning customers are cheap to win and often would have bought anyway. Growth comes from new ones. Google Ads lets you set customer acquisition goals so new customers are valued more or targeted exclusively, but it needs a customer list and a clear view of what a new customer is worth over time. An agency that knows DTC treats this as a decision, not a checkbox.
Ask: "How much more can we afford to pay for a new customer than a returning one, and how would you set that up?"
Worry if: they have not heard of new customer acquisition goals, or switch them on without a number behind them.
6. Measurement you can trust
If tracking is off, every decision above it is off too. That means conversion tracking checked against real orders, enhanced conversions, consent mode handled properly, and for Shopify brands a setup that survived the recent checkout changes (our Shopify conversion tracking guide covers the details). Beyond the platform, a serious agency also looks at blended numbers: total revenue against total ad spend across channels, because Google will happily take credit for sales that Meta or email started.
Ask: "Before you change anything, how will you check that the conversions in Google Ads match our actual orders?"
Worry if: "We trust the platform data." No you do not, not until it has been reconciled.
7. A steady creative supply
Performance Max, Demand Gen and YouTube all need images, videos and copy, and they need fresh ones. Assets wear out. An agency that only manages bids and keywords will leave these campaigns running on the same five images for a year. That does not mean the agency has to shoot your content, but it should tell you exactly what it needs, in what formats, and how often, and it should read asset performance to decide what to make next.
Ask: "What creative will you need from us each month, and how do you decide what to replace?"
Worry if: "Google generates the assets automatically." It will, and you will not like them.
8. Scaling in controlled steps
Doubling a budget overnight resets learning and usually tanks efficiency for weeks. Good scaling is boring: budget raised in steps of around 20% when the numbers support it, a week or so between changes, targets moved gradually, and a clear rule for when to stop. A USA holster brand more than doubled revenue to $400K+ a month at 5.45x ROAS this way, through structure and patience rather than one big jump.
Ask: "If we wanted to double spend in the next quarter, what would the first six weeks look like?"
Worry if: "Just raise the budget and we will see what happens."
9. Reporting and ownership
You should own the Google Ads account, Merchant Center and GA4, with the agency added as a user. Reporting should say what changed, why, and what it did to profit, in plain language. A dashboard link with no commentary is not reporting.
Ask: "Whose name is on the account, and what happens to our data if we part ways?"
Worry if: they want to run the account inside their own manager account with you as a guest, or the exit terms are vague.