You have the product, the site is live, but you only get three orders a week. A familiar picture for most new e-commerce projects. PPC advertising for online stores solves this problem faster than any other channel: a person is already searching to “buy an Indesit ITS 4200 W refrigerator” — and sees your product with the price in the Google search results. Not in six months, as with SEO, but a couple of days after launch.
However, there is a catch. In e-commerce, everything is more complex than in services. There isn’t just one landing page here, but hundreds or thousands of product pages, each with its own price, availability, and profit margin. Therefore, the approach is different: Shopping campaigns instead of text ads, a product feed instead of manual ads, ROAS instead of cost per lead (CPA). Let’s break it all down step by step — from preparing your catalog to calculating the budget and typical mistakes.
How does PPC advertising for an online store differ from services?
A store sells an assortment of products, not just a single offer. Almost everything else stems from this.
- First, you need a lot of ads. A lawyer can get by with a dozen keywords and one landing page. An electronics store with 3,000 items cannot do this — you can’t manage that volume manually. Google Ads solves this issue through a product feed: a data file from which the system builds the ads itself.
- Second, you have to count money, not leads. An inquiry for an apartment renovation and an order for a 199 UAH phone case are different types of requests, even though both are technically conversions. For a store, the main metric is ROAS (Return on Ad Spend): how much revenue each invested hryvnia brings in. A ROAS of 400% means that 1,000 UAH spent on advertising brought in 4,000 UAH in revenue.
- Third, dynamics are important here. If a product is out of stock, the ad should disappear. If the price drops, it must be updated in the ad. Manual management cannot handle this; automation is required.
Hence a simple conclusion: PPC advertising for an online store is built around product data, not around a list of keywords. Keywords are also needed, but they supplement it rather than replace it.
Which campaigns should a new online store launch first?
This is the question that almost every consultation starts with. I’ll answer right away: Google Shopping, then search and remarketing. Performance Max — after statistics have accumulated. Now let’s look at each type in more detail.
Google Shopping (product advertising)
These are the product cards with a photo, price, and seller’s name above the search results. In the interface, they are called Google Shopping. For e-commerce, this is number one, and here’s why:
- the user sees the price before clicking, so they are already “warm” when they visit;
- ads are created automatically from the feed — no need to write ad copy;
- the conversion rate is on average higher than that of text ads, and the cost per click is cheaper in most niches;
- ads are shown based on actual search queries without manually compiling keyword lists.
Let’s say a children’s shoe store in Dnipro uploaded a feed with 800 items. Two days later, ads are running for the queries “children’s fall boots,” “girls’ sneakers size 30” — even though the owner didn’t enter a single keyword. Google matched all this with the titles and descriptions from the catalog on its own.
The flip side: the quality of impressions directly depends on the feed. The title “Boots 123” won’t attract anyone. The title “Geox kids’ demi-season boots, leather, size 30” will.
Search campaigns
Classic text ads based on keywords. For a store, they cover three tasks:
- Branded queries. “Rozetka laptop” goes to Rozetka, and “buy laptop in Your-Brand online store” goes to you. It’s cheap and protects you from competitors.
- Category queries. “Buy Samsung smartphone” is a general query; it’s better to direct this to a category page rather than a specific product page.
- Queries with specific qualifiers. “Delonghi coffee machine with delivery in Ukraine,” “winter tires 205/55 r16 on installment” — Shopping ads don’t always process such qualifiers well, but text ads will.
Dynamic Search Ads (DSA) also come in handy here. In this case, Google crawls the site’s content and automatically matches a headline to the query. For a catalog with thousands of pages, DSAs save a lot of manual work and simultaneously suggest what should be separated into specific ad groups.
Performance Max
PMax combines Search, Shopping, GDN, YouTube, Gmail, and Discover. There are few settings: goal, budget, feed, a set of images and texts, and audience signals. The algorithm then decides on its own where and to whom to show the ads.
It sounds convenient, and in mature accounts, the results are good. But for a new store without a conversion history, PMax is usually harmful: the algorithm has nothing to learn from, and for the first two or three weeks, the budget is wasted on chaotic impressions across the GDN and YouTube. A practical rule of thumb is to enable Performance Max when the account steadily brings in 30–50 orders per month through Google Shopping and Search.
Attention. Performance Max eagerly picks up branded queries and takes credit for them. The ROAS in the report looks great, but the actual increase in sales is small. Exclude your brand from PMax using negative keyword lists at the campaign level to see the true picture.
Remarketing and GDN
The Google Display Network consists of banners on Google’s partner websites. As an independent channel for attracting “cold” customers, it is poorly suited for stores: people click a lot, but buy rarely. However, as a platform for remarketing, it is indispensable.
Standard remarketing shows all past visitors the same banner: “Come back, we have discounts.” It has an effect, but a weak one. A store needs dynamic remarketing — it shows a person the exact item they were looking at, with the current price from the feed.
The most profitable segment is abandoned carts. A person added headphones for 2,500 UAH to their cart, reached the checkout, and closed the tab. There are plenty of reasons: they got a call, got distracted, or decided to compare prices on Prom. A banner with those same headphones a few hours later, or even better, with a 5% promo code or free shipping, brings back 5 to 15% of these people. For comparison, the conversion rate of cold traffic from the GDN is rarely above 0.3%.
What is needed for setup:
- A Google Ads tag with product parameters (id, price, page type) on all site pages — most CMS platforms install it via a built-in module or Tag Manager.
- The same feed from Merchant Center — images and prices will be pulled from it.
- Audiences: “viewed item, didn’t add to cart” (30 days), “added to cart, didn’t buy” (7–14 days), “purchased” (for exclusion or upselling in 30–60 days).
Bids vary by segment. An abandoned cart is worth the most, a product page view is worth less, and buyers are excluded from the main campaigns so they aren’t chased with items they’ve already bought.
Look-alike audiences are also built based on buyers (in Google Ads they are called “similar segments” and work within PMax and the GDN). The system looks for people whose behavior resembles that of your customers. With 500+ orders, this is a cheap way to expand your reach without drifting into entirely cold audiences.
What should an online store prepare before launching ads?
This is where people most often save time, and then lose money. Ads bring people to the site — if it isn’t ready, they leave, but the money for the click has already been spent. Go through the list in advance.
- Product page. A photo on a white background (this is a system requirement, not a suggestion), price with no surprises, availability, delivery time, and a “Buy” button visible above the fold on mobile. Check it on a real smartphone, not in a browser emulator.
- Checkout. Count the fields in the form. Eight fields are too many; three or four are fine. Add a guest checkout option and cash on delivery via Nova Poshta: for a Ukrainian, this is still a compelling argument.
- Analytics. Install Google Analytics 4 and set up purchase tracking to pass the transaction value. Without this, you can’t calculate ROAS, meaning you won’t know which campaigns feed the store and which eat the budget. Import the “Purchase” conversion into Google Ads — automated strategies will use it to learn.
- Merchant Center and product feed. A Merchant Center account is mandatory for launching Shopping and PMax campaigns. A list of items with required fields (id, title, description, link, image, price, availability, brand, GTIN barcode, condition) is uploaded there. Most Ukrainian CMS and website builders (Horoshop, Prom, OpenCart, WooCommerce) can generate such a file automatically. Ensure the price in the feed matches the site—a mismatch leads to product disapproval, and sometimes even account suspension.
- Filled-out shipping and return policies. If you don’t provide them, the catalog won’t pass moderation. Fill them out honestly: if returns are legally 14 days, just state that.
- Usability and speed. A page that takes longer than 3–4 seconds to load on a mobile phone loses up to half of its visitors. Compress images, remove unnecessary scripts, and test it via PageSpeed Insights.
Mobile traffic: a separate conversation
In Ukrainian e-commerce, 70–80% of clicks on PPC advertising for online stores come from mobile phones. However, the mobile conversion rate is usually one and a half to two times lower than on desktop. This means that two-thirds of the budget is spent on devices where people are less likely to buy.
Don’t disable mobile impressions — you’ll lose most of your demand. Instead:
- verify that the price and the “Buy” button are visible without scrolling;
- shorten the checkout form to a minimum, and enable Nova Poshta address auto-fill;
- add Apple Pay / Google Pay — entering a card number on a phone is inconvenient, and some people won’t complete the process;
- set up device bid adjustments if you are using manual bidding strategies: minus 20–30% for smartphones if sales are noticeably worse there.
Look at the device report in GA4. Sometimes it turns out that people buy just as well, if not better, from a phone — just at a different time of day.
Step-by-step setup of PPC advertising for an online store
Below is a sensible workflow if you are launching an account from scratch. Setting up PPC advertising for an online store takes anywhere from three days to two weeks, depending on the size of the catalog.
- Create a Google Ads account and link it with GA4 and Merchant Center. Choose the hryvnia as your currency — you can’t change it later.
- Set up conversions. The primary one is “Purchase” with a value. Secondary ones are “Add to cart” and “Begin checkout” — do not include these in optimization, just use them for observation.
- Upload the feed to Merchant Center and wait for moderation (usually 1–3 days). Fix anything that gets disapproved.
- Launch a Standard Shopping campaign. Divide products into groups by category or margin — it makes bid management easier. For the first two weeks, use Manual CPC or “Maximize clicks” with a bid limit.
- Compile your keyword lists: brand, categories, best-selling models. Add negative keywords right away — “free,” “used,” “DIY,” “reviews,” “manual,” and names of cities you don’t deliver to.
- Launch Search campaigns. One for the brand, one or two for categories. Do not mix them: branded traffic has a conversion rate 5–10 times higher, which will skew the statistics.
- Enable DSAs for the rest of the catalog — they will pick up what you missed.
- Set up GDN remarketing targeting visitors from the last 30 days.
- In 3–4 weeks, when you’ve gathered at least 30 purchases, switch Shopping and Search to automated bidding strategies — “Target ROAS” (tROAS) or “Maximize conversion value.”
- Another month later, test Performance Max with a feed, excluding branded queries.
The workflow above isn’t set in stone. If your budget is very small (say, 10,000 UAH per month), launch only Shopping and Brand campaigns — and don’t spread yourself too thin.
Instructions for setting up advertising for online stores on different CMS
Although the general principle of setting up contextual advertising for online stores on different CMSs is approximately the same, there are still some nuances that may differ depending on the specific platform. Our Google logistician, Yana Lyashenko, sorted them out in detail.
Setting up advertising
for Horoshop Articles on
similar topics
Setting up ads
for Shopify Articles on
similar topics
Setting up ads
on Prom.ua Articles on
similar topics
What budget is needed to launch advertising for an online store?
There is no one-size-fits-all number — it all comes down to the niche, profit margin, and cost per click. But you can calculate it, and it’s better to do this before launch rather than being surprised a month later.
Let’s look at an example. A coffee and accessories store, the average order value is 1,200 UAH, the margin is 35%, meaning 420 UAH remains from one sale. The average cost per click in Google Shopping for this niche is about 6 UAH, and the site’s conversion rate is 1.5%.
Let’s calculate:
- one order requires about 67 clicks (100 / 1.5);
- 67 clicks × 6 UAH = 402 UAH — this is how much the order costs to acquire;
- the 420 UAH margin minus 402 UAH = 18 UAH profit per order.
Barely breaking even. With these inputs, the advertising pays for itself, but it doesn’t bring in money. What can be done? Either increase the conversion rate (from 1.5% to 2.5% — and the cost per acquisition drops to 240 UAH), or increase the average order value through bundles and upselling, or lower the CPC by creating more precise groups in the feed. Usually, all three levers are pulled at once.
The minimum ROAS to break even is calculated as 100% / profit margin. With a 35% margin, this is 286%: every hryvnia spent on advertising must return 2.86 UAH in revenue. Anything below this, and the store is operating at a loss, even if there are many orders.
Now about the starting amount. For the test, you need to gather statistics: at least 30–50 orders in a month for automated bidding to kick in. In the example above, that’s 40 × 402 ≈ 16,000 UAH plus Search and Remarketing — totaling 20,000–25,000 UAH for the first month. For niches with expensive clicks (electronics, furniture, auto parts), the sum rises to 40,000–60,000 UAH; for inexpensive assortments with a high conversion rate, it drops to 8,000–12,000 UAH.
Attention. The first month’s budget is the price you pay for data, not for profit. PPC advertising for an online store rarely pays off in the first week. Expect the ROAS to be below the target for the first 3–4 weeks. If you quit everything in the second week because there are no sales, the money will be wasted, and no conclusions will be drawn.
A separate expense item is campaign management. Agencies in Ukraine charge anywhere from 8,000–10,000 UAH per month for a small store to 50,000 UAH if the assortment is very large. Doing it yourself is also feasible, but budget 5–8 hours a week for negative keywords, bids, and reports — at least for the first couple of months.
Common mistakes when setting up Google Ads for an online store
Over years of working with e-commerce projects, the exact same set of problems repeats with enviable regularity. Before you launch PPC advertising for an online store, check yourself.
- No tracking of the order value. Conversions are counted, but they are all equal to one. A campaign with ten orders at 300 UAH looks better than one with three orders at 5,000 UAH. In this situation, automated bidding optimizes in the wrong direction.
- One feed for everything, without segmentation. 2,000 product pages in one ad group with a single bid. Bestsellers and dead stock receive the same amount of money. Break them down at least by category, or better yet, by margin and inventory turnover.
- PMax on day one. We’ve already discussed this. The algorithm needs data, otherwise it burns the budget on YouTube.
- Mixing Search and GDN. When creating a Search campaign, Google enables GDN impressions by default. Uncheck that box. Otherwise, half of your “search” budget will go to banners in mobile games.
- Not collecting negative keywords. Once a week, open the Search Terms report. You are bound to find “download,” “DIY,” “photo,” “which is better” — all of this was paid for and brought you nothing.
- Products in the feed without a GTIN and brand. Google demotes them in Shopping campaigns or disapproves them altogether. Fill in the barcodes — they are on the packaging or on the supplier’s invoice.
- Ads leading to the homepage. For the query “buy Redmond RMC-M90 multicooker,” a person should land on the RMC-M90 product page, not the homepage featuring a banner about an iron promotion. Every extra click means minus 20–30% to the conversion rate.
- Not excluding buyers from remarketing. A person bought a refrigerator, and for the next three weeks, refrigerators haunt them. It’s annoying and wastes money.
- Stopping ads in the second week. “We wasted 5,000 UAH, there are no orders, so it doesn’t work.” Five thousand UAH means 800 clicks. At a 1.5% conversion rate, 12 orders are expected, but the variance on such a small sample is huge. Only draw conclusions based on at least a full month.
Cases on setting up advertising for online stores
Hyro scooters, segways
Mini Segway, Esway
Fireplaces & Stoves
Fireplace Accessories
Stationery
Online stationery store
Which metrics to watch after launch?
Not all numbers in Google Ads are equally useful. When launching PPC advertising for an online store, high CTRs and ad positions are nice, but they don’t directly affect profits. Look at four things:
- ROAS by campaigns and product groups — this is your main compass. Compare it to your break-even point, not “to last week.”
- Cost per Acquisition (CPA) — handy for understanding how much money it takes to acquire one customer.
- Site conversion rate by source. If Google Shopping has 2% and Search has 0.6%, the issue lies with the landing pages, not the ads.
- Search and Shopping impression share. If it’s below 30%, you are rarely seen; there is room to grow your budget or bids.
