Modern Paid Search Marketing Strategies For The Ai Age

In this 6 part essay I'll look at the how paid search has evolved over the last 5 years (2020 - 2025), explaining why old metrics and structures now fall short, and how to re-engineer your campaigns for true incrementality, new customer growth, and profit in the age of ai.

Beware of “rosy” platform metrics. One of the first wake-up calls is that Google Ads’ own attribution often inflates your reported ROAS. Ad platforms commonly count conversions that would have happened anyway (organically or via another channel) as if the ads caused them​. For example, if a shopper sees a Facebook ad in the afternoon and then searches Google for your brand in the evening to buy, both Facebook and Google will each claim the full revenue credit​. Neither platform knows about the other, so each reports a high ROAS, even though together they’re double-counting the same sale. The result? You might celebrate a 500% ROAS in Google Ads, not realizing a chunk of those conversions were coming to you regardless. Data-driven attribution (Google’s default model now) distributes credit across touchpoints better than last-click did, but it still doesn’t tell you how many sales only happened because of your ads. The key is to measure incrementality, the true causal impact of ads on revenue. Experienced marketers now run geo holdout tests or use Conversion Lift experiments to find out how much of Google’s reported conversions are incremental. Often, it’s eye-opening: one agency’s geo tests found that certain brand campaigns generated “no additional (net) revenue”, 0% incrementality, even though Google credited them with many conversions​. In short, don’t take platform-reported ROI at face value; back it with experiments and marketing mix modeling (MMM) to get the real picture.

Google’s automation has changed the game, and not always in your favor. Another reality check: the latest Google Ads products (like Performance Max campaigns and broad match keywords with Smart Bidding) can drive great results, but they also blur the lines between brand and non-brand, prospecting and retargeting. If used naively, they can mislead you on incrementality. For instance, Performance Max (PMax) is Google’s AI-driven, all-in-one campaign type that replaced Smart Shopping. By design, PMax will try to capture any conversion it can. It’ll serve shopping ads, search ads, display, YouTube, whatever gets a conversion. The problem is PMax tends to cannibalize easy wins from your existing traffic: it will eagerly serve ads on your branded terms or re-engage past site visitors if that scores a conversion​. Those conversions look great on paper, but they’re often customers who were already familiar with you. Google even claims that PMax’s brand-driven conversions are incremental, but many brands have found the opposite. In fact, one PPC agency reported that keeping branded terms inside PMax “gives Google a license to waste your money” by inflating conversion counts with people who would have converted anyway. PMax also blends in retargeting (e.g. showing display ads to past customers) with no way to separate it​, which artificially boosts ROAS, the campaign looks amazing because it’s counting repeat buyers as “conversions” it drove. The takeaway: Google’s AI isn’t magically delivering only new customers; it’s often cherry-picking low-hanging fruit. As advertisers, we need to guide the automation (through proper exclusions and structure, discussed later) to ensure we’re truly growing business, not just reallocating existing demand to paid channels.

Search volume limits your growth, you can’t spend beyond the demand that exists. In 2020, many brands believed if you increase your Google Ads budget, conversions will scale linearly. In 2025 we know better: paid search is ultimately bounded by how many people are searching for your product or category. You can’t force more people to search on Google, if you already dominate the search results for your keywords, doubling your budget won’t double your sales. Marketers often hit a plateau where campaigns become “Limited by search volume.” One SaaS marketing study noted that after you’ve harvested the available low-volume, high-intent keywords, pouring more money into them yields diminishing returns​. 

You might see impressions and clicks max out, or Google Ads will even warn you a campaign is “Limited by search volume.” At that point, continuing to raise bids or budgets just means you’ll pay more per click (or start showing on tangential queries) without a commensurate increase in conversions. To break through this ceiling, you need to create or capture new demand beyond existing search queries. This can mean broadening your keywords cautiously (e.g. using broad match with smart bidding to find new relevant queries), expanding to new channels (like Bing Ads, which can tap a different audience), or moving up the funnel with content to generate more interest (more on that in the Demand Gen section). The key is to recognize when you’ve saturated the core search terms. Also, distinguish brand search volume vs. generic. Branded search campaigns are almost naturally capped by how many people search your brand name. As one incrementality expert put it, “brand keyword campaigns are almost naturally capped by the amount of brand keyword searches”. 

If your brand isn’t widely known, there’s only so much your brand campaign can spend, and that’s a good thing! Don’t try to force more spend there; instead focus on growing the brand searches themselves via PR, offline marketing, or upper-funnel ads. In summary, appreciate that demand is finite on search. Hitting a high impression share (we’ll discuss this metric later) is a sign you’re near the cap. From there, smart marketers pivot to new strategies rather than overspending for minimal gains.

Structured campaigns are your ally for true incrementality. A final reality check: achieving incremental growth from Google Ads isn’t just about budgets and bids, it’s about how you structure your account to guide Google’s automation. In the past, we obsessed over tactics like SKAGs (single-keyword ad groups) for control. Today, structure serves a different purpose: separating traffic by intent and value so you can apply the right strategies to each. If you lump everything into one smart campaign, you’ll never know if your spend is actually bringing in new business or just capturing existing demand. For example, running a combined brand + non-brand campaign makes it impossible to measure the true performance of your non-brand ads, the stellar ROI of brand keywords will mask the lower ROI of prospecting keywords​. 

Similarly, mixing new and returning customers in one campaign makes it hard to see if you’re expanding your customer base or just re-selling to the same folks. The solution is proper campaign segmentation: isolate brand traffic in its own campaigns, segment new vs. existing customers where possible, and use separate budgets/goals for different product categories or audiences. By structuring this way, you create natural “experimental control groups” in your account. You can clearly compare, say, brand vs. non-brand performance, and allocate spend to where it drives net-new value. A well-structured account also prevents Google’s AI from unintentionally cannibalizing one area to feed another. For instance, if you exclude your brand terms from a Performance Max campaign, it forces PMax to work harder to find new customers instead of just leaning on cheap brand clicks. As we’ll cover next, the core account structure for 2025 deliberately segments campaigns in a way that aligns with incrementality and business goals. In 2020 you might have let Google automate everything in one bucket; in 2025, you let it run free within guardrails that you set via smart structure.

A modern Google Ads account in 2025 looks very different from the typical 2020 setup. We now design structures that play to Google’s automation strengths while maintaining transparency and control. The goal is to let machine learning optimize within well-defined swimlanes. Below is a blueprint of an optimal account structure and spend allocation for an e-commerce brand (percentages can be adjusted for your business). This structure has three primary pillars, each with a distinct role:

Primary Driver (50–70% of spend), Feed-Driven Shopping Campaigns (Non-Brand)

For most brands, the workhorse of Google Ads in 2025 is a product feed-driven campaign that captures high-intent shoppers. This is typically done with Standard Shopping or a Performance Max campaign without assets (also known as a feed-only PMax). By not adding any headline/text/image assets to PMax, you essentially restrict it to act like a Smart Shopping campaign, serving Shopping ads on Search and the Display Network using your product feed. Why this approach? It harnesses Google’s latest automation for bidding and query matching, but keeps the focus on bottom-of-funnel buyers looking for your products, rather than distracting the campaign with creative-heavy prospecting.

• Segment by product category or type. Rather than one giant shopping campaign for all products, break this Primary Driver into multiple campaigns or Performance Max asset groups divided by category, product line, or profitability tier. Category-level segmentation ensures each campaign has a cohesive product set and sufficient data to let the algorithm learn. It also lets you apply different bidding strategies per category, for example, you might set a higher Target ROAS on low-margin electronics and a lower Target ROAS (higher spend aggression) on high-margin accessories. If a category is too small (limited volume), group it with related items so the campaign isn’t starved for data​. The structure might be: Campaign 1 - Category A, Campaign 2 - Category B, etc., each targeting its own product group. This prevents a scenario where one bestseller item soaks up all the budget at expense of others; instead, spend is balanced and aligned to your business priorities.
• Use Smart Bidding with ROAS targets by category. All Primary Driver campaigns should use Target ROAS bidding (tROAS) or Maximize Conversion Value with a ROAS goal, set according to what each category can tolerate. In 2025, manual bidding is virtually obsolete for shopping ads, Google’s machine learning can adjust bids per auction far better than any human. Your job is to feed the right goal. Calculate realistic ROAS targets based on your margins and desired volume. For instance, if Category A has 50% gross margin, you might aim for a 2x ROAS (50% cost/sales) to break even, or higher for profit; if Category B has 70% margin, you can profit at a lower ROAS so you might set 1.5x to aggressively grow it. These targets can be input as campaign-level Target ROAS. Monitor actual performance and calibrate over time. The key is not using one-size-fits-all goals. By tailoring ROAS targets, you tell Google’s bidding algo where to be aggressive and where to be conservative, which maximizes profit. (Advanced tip: if you have precise profit data, consider using Google’s new Profit Optimization feature or value rules to bid to profit instead of revenue, more on profit metrics later.)
• Exclude existing customers to focus on new ones. A huge advancement by 2025 is the ability to optimize for new customer acquisition (NCA) directly in Google Ads. In your Primary Driver shopping campaigns, always enable new customer settings. For Performance Max, you can set it to “New Customer Only” mode or apply a value uplift for new customers. In New Customer Only mode, Google will exclude your existing customers from bidding and only go after truly new users​. (This relies on you uploading your customer list to Google Ads, so it recognizes who’s existing.) The rationale is simple: you don’t need to pay Google ad dollars to get repeat purchases from loyal customers, those you can reach via email, SMS, or they’ll come direct. Your precious ad budget here should go to acquiring net-new buyers. By excluding known customers, you immediately improve the incrementality of these campaigns. If completely excluding is too strict (e.g. you also want some retargeting), another approach is “bid higher on new customers”, Google allows adding an extra conversion value or target adjustment for new vs. returning. Either way, make NCA a mandatory setting. Brands that neglect this are often shocked to find their shopping campaigns were largely retargeting their own customer base (and thus the true incremental ROAS was much lower than reported). As Optmyzr’s experts note, not defining new vs. existing can result in “overpaying for existing customers, thus distorting ROAS”​. We want these primary campaigns laser-focused on people who haven’t bought from us before.
• Keep brand terms out to measure true performance. This point cannot be stressed enough: exclude or separate brand traffic from your Primary Driver campaigns. If you’re using Standard Shopping, you can’t directly add keywords, but you can use campaign priority settings and negative keywords to prevent your shopping ads from triggering on your own brand name. If using PMax, Google now (finally) allows adding a list of negative keywords or a brand exclusion signal, take advantage of that to block brand searches. The reason is that brand queries (e.g. someone searching your exact product name or your store name) will naturally convert at a very high rate, they likely already intended to buy from you. Including those in the primary campaign will inflate its ROAS and make it seem like it’s hitting goal easily, when in fact it’s just capturing revenue you’d have gotten organically. One agency’s analysis of PMax found that including brand hides poor performance in other areas: e.g. just 10% spend on brand at 20x ROAS can make the other 90% of spend at 0.5x ROAS look like a decent 2.5x average​. That’s dangerous for decision making. By excluding brand, you force your primary campaigns to stand on their own merit in driving non-brand sales. Your reported ROAS will be lower, but far more reflective of true incremental performance. (Don’t worry, we will still cover brand defense in its own campaign next.) As PPC Hero notes, Google makes it challenging to remove brand from PMax and claims those brand conversions are incremental, but agencies have found time and again that keeping branded terms out leads to more efficient spend​. The bottom line: keep your Primary Driver shopping campaigns focused on generic and competitor searches, not your easy brand traffic.
• Leverage feed optimizations instead of ad copy. In shopping campaigns, you don’t have keywords or ad text to control search targeting, Google matches based on your product feed info. Therefore, optimizing your product feed is a critical execution detail (expanded in Section 5). Use descriptive, keyword-rich product titles and correct product types/categories so that Google understands when to show your products. For example, if you sell running shoes, ensure the title contains brand, model, “running shoes”, color, etc. This will help your products serve on more relevant queries. Also, use Custom Labels to tag products by attributes like margin, seasonality, or price tier, this allows you to break out high-value vs. low-value items into different campaigns or bid strategies. The feed-only PMax will use these signals to group products into listing groups. Essentially, think of feed-driven campaigns as the new keyword: your feed content and campaign structure is how you target. A well-structured feed (with all GTINs, MPNs, correct Google Product Categories, etc.) will gain more impressions and cheaper CPCs than a poor feed. We’ll dive deeper into feed management later, but suffice to say that the Primary Driver depends on a healthy, optimized product feed fueling it.

In summary, the Primary Driver campaigns act as your core revenue engine, capturing demand from people actively searching for products you sell (excluding pure brand searches). By segmenting by category, using smart bidding with ROAS targets, focusing on new customers, and excluding brand, you create a machine that reliably turns ad spend into incremental sales. For many brands this accounts for roughly 50–70% of the total Google Ads budget, as it should, it’s where the lowest funnel, most ready-to-buy users are. Now, let’s talk about the other pieces that complement this engine.

Brand Search Defense (10–20% of spend), Protect Your Turf with Branded Keywords

Even if brand campaigns don’t drive “new” demand, almost every brand still needs to run them. Brand search defense ensures that when someone searches your brand or product names, they find you first, not a competitor’s ad. It also helps capture high-intent users who include your brand in broader queries (e.g. “<YourBrand> reviews” or “buy <YourBrand> Model X”). In 2025, the approach to brand campaigns has evolved to be efficient and simple:

• Use broad match for your brand keywords. This might sound counterintuitive to old-school PPCers who feared broad match, but today broad match with smart bidding is very effective, especially for brand terms. By adding a broad match keyword of your brand name (and perhaps product trademarks), Google will automatically cover essentially all variations, misspellings, and long-tail searches that include your brand. For example, if your brand is “Acme Widgets”, a broad match acmewidgetsacme widgetsacmewidgets can match queries like “acme widget pricing”, “buy acme gadgets” (even if misspelled), or “acme widget model 123 review”. This saves you from having to enumerate every possible brand term. Broad match has become the default match type in Google Ads as of 2024​, thanks to improvements in Google’s AI understanding of intent. Google’s own data shows that broad match, when paired with Smart Bidding and relevant ads, can reach more relevant searches within your goals​. For brand defense, this is perfect: you want maximal coverage of any search containing your brand. You generally do not need any negative keywords in a pure brand campaign, because almost any query containing your brand is something you want to show up for. (One exception: if your brand name is a common word or overlaps with unrelated meanings, you might add negatives for those cases. But for most distinct brand names, no exclusions are needed.)
• Keep it simple, one campaign, few keywords. Unlike the Primary Driver, you do not need granular segmentation in brand campaigns. Often a single campaign with one ad group is sufficient, containing a broad match for the brand and perhaps an exact match for the brand if you want to ensure coverage (though Google will match the broad to exact queries anyway). You’re not trying to “optimize” across many keywords; the goal is blanket coverage. Structure isn’t complex here because the intent is all the same (navigational or high intent for your brand). Also, you typically won’t use audience segmentation or time-of-day nuances, brand demand is captured whenever it occurs. This simplicity makes it easier to manage and to analyze. It also means you can apply a straightforward bidding strategy and budget without much micromanagement.
• Set a high efficiency goal (or manual bidding) for maximum profit. Since branded searches convert extremely well (it’s not uncommon to see 30-50% conversion rates on pure brand terms) and have low CPCs (quality score is high and competition is low for your own brand), the ROAS on brand campaigns tends to be very high. You want to keep it that way. One strategy is to use Target ROAS bidding and set an aggressively high ROAS target (e.g. 1000% or more, depending on your numbers) so that Google’s algorithm spends only what’s needed to capture the core brand traffic. Essentially, you’re telling it “don’t chase marginal brand impressions if they become too expensive; only give me the cream of the crop.” In practice, even a high target ROAS will still let your brand ad dominate because the bids required are low, but it prevents Google from overspending if, say, someone types a very broad query that mistakenly matches. Another approach some experts take is to use manual bidding or Enhanced CPC for brand campaigns to more directly control bids. The logic: you want near 100% impression share for your brand terms (so that a competitor can’t steal those clicks)​, and manual bidding makes it easy to just bid high enough to always be on top. Brand clicks are cheap, so even bidding to position 1.0 constantly won’t break the bank. The downside of manual is you lose the automated optimizations, but for brand it’s usually fine. Whether you use tROAS or manual, the guiding principle is efficiency, aim for the maximum share of voice on your brand at the lowest practical cost. You might find your brand campaign is getting, say, a 10:1 ROAS (i.e. 1000% ROAS) or more. If it’s much lower, check if non-brand queries are accidentally matching or if competitors are driving up your CPC (in which case, you may need to bid a bit more to outrank them). Many companies actually see brand campaigns as pure profit drivers that also support other channels (since people often search brand after seeing an ad elsewhere). So treat it as a defensive, high-ROI bucket, not a growth engine.
• No need for fancy creative, focus on links and extensions. Someone searching your brand already knows you; the job of your brand ad is simply to direct them to the right place efficiently. Make sure you use sitelink extensions abundantly on brand campaigns. Sitelinks can guide users to specific high-interest sections (e.g. “New Arrivals”, “Support”, “Contact Us”, “Deals”, etc.) directly from the ad. This not only improves user experience but also boosts your ad’s CTR. In fact, simply adding sitelink extensions has been shown to give about an 8% uplift in CTR on ads. A higher CTR on your brand ad means fewer people clicking down to see competitors. Ensure your ad copy highlights your brand strengths (“Official Site”, free shipping, etc.), but you don’t need to oversell, the user is already convinced enough to search your brand. Also, unlike 2020 where we might separate brand keywords by match type, nowadays one broad match keyword with smart bidding and a strong responsive search ad will do the trick. Keep ads up-to-date though: if you’re running a big promotion or have a new tagline, reflect it in your brand ad copy since these users are very valuable. All in all, the brand search defense layer is relatively low maintenance, but it’s crucial to have in place to ensure competitors or resellers aren’t poaching your customers via search ads. Allocate about 10–20% of spend here (sometimes even less, depending on how big your brand volume is). If this starts creeping higher, it might be a sign you’re including too many non-brand terms or that your brand popularity has surged (good problem to have, you might then lower the % on other areas).

Demand Gen & YouTube (10–20% of spend), Fueling the Upper Funnel and Creating Intent

The third pillar of the 2025 account structure goes beyond capturing existing demand, it’s about generating new demand and guiding it to conversion. Google’s ecosystem now offers powerful tools for what used to be called “Display/Discovery” advertising, as well as YouTube for video. In particular, Google introduced Demand Gen campaigns (replacing Discovery campaigns) which run across YouTube Home feed, YouTube Shorts, Gmail, and Discover feed, basically Google’s answer to social media style ads. Alongside, YouTube Ads (especially YouTube TrueView and Video Action campaigns) allow for rich storytelling and product education. We allocate ~10–20% of budget here to feed the top and mid funnel, knowing that this spend doesn’t always convert immediately in-platform, but creates lift that helps the search and shopping campaigns. Here’s how to execute this layer effectively in 2025:

• Differentiate high-intent content from pure awareness. Not all “upper funnel” campaigns are equal. Some content is still conversion-oriented even if it’s not search, for example, a YouTube review or tutorial video about your product can capture users who are close to purchase (they’re actively researching). Contrast that with a generic lifestyle video ad that’s just to build brand familiarity (true top-funnel awareness). Both have a place, but you want to separate these strategies. High-intent demand gen means using content/topics that people might actively seek out or be responsive to shortly before buying. These include things like “How to use <Product>”, “<Brand> vs Competitor comparison”, “<Industry> tips and tricks” where your product naturally fits as a solution. You can reach such users on YouTube by using Custom Intent audiences (people who recently searched certain keywords can be targeted with video ads), for instance, target viewers who searched for your competitor’s name, or for “[Your product] review”. According to a PPC practitioner on Reddit, using custom intent audiences for people who searched competitor domains is one of the best ways to start driving conversions on YouTube​. On the other hand, top-funnel awareness might target broad Affinity audiences or demo targeting with a inspirational ad that isn’t expected to convert immediately. Why distinguish? Because you measure and optimize them differently. High-intent video campaigns can be optimized for conversions or set with tCPA goals, whereas pure awareness you might optimize for views or engaged reach. In your account, consider having at least two separate campaigns: e.g. “YouTube, MOFU (tutorials/reviews)” and “YouTube, TOFU (brand intro)”. Similarly, for Demand Gen (Discovery ads), you might have one campaign targeting in-market audiences with product-focused imagery and another targeting broad lifestyle interests with soft-sell content. This separation ensures the budget in each is used for its specific purpose and you can evaluate performance appropriately (conversion impact vs. reach/engagement).
• Leverage tutorials, how-tos, and reviews as creative, they can convert! A major advanced strategy is using educational content to drive conversions on Google’s video and discovery channels. For example, create a 2-3 minute tutorial video or product demo that addresses a common problem and shows your product as the solution. Promote this video via YouTube TrueView for Action campaigns (which include a prominent CTA button). Users who watch such content are highly qualified, they’re essentially pre-selling themselves. Best practices here include: start the video by identifying a pain point your target customer has (hook them in), then demonstrate how your product or service solves it (with clear visuals, possibly testimonials), and end with a direct call-to-action (“Get yours now” or “Learn more on our site”). This format tends to weed out uninterested viewers and engage the ones who matter, often leading them to click through. Another great format is a comparison or review video, either one you produce or working with influencers/creators. Someone searching for “Best wireless earbuds 2025” on YouTube is likely deciding what to buy; if you show up with an ad or sponsored video reviewing top options (and featuring your brand), you can influence that decision. The key is the video must be genuinely helpful and not just a pure ad, otherwise users will skip. When done right, YouTube ads are not “just for awareness”, they absolutely can drive conversions (contrary to what many thought in 2020). One advertiser reported significantly growing sales with a $27 online course by fine-tuning YouTube campaigns over 15 months, proving that with the right targeting and content, YouTube can produce direct ROI​.
• Target wisely and use Google’s audience data. On the execution side for Demand Gen/YouTube, 2025 Google Ads offers rich audience targeting options. Take advantage of Custom Segments (formerly custom intent/custom affinity) where you input keywords, URLs, or apps that your ideal customer is interested in. For example, for a home workout equipment brand, you might target a custom segment of people who searched for “home workout ideas” or visited fitness competitor sites, these can be used in Demand Gen campaigns to show them engaging image ads of your equipment in use. For YouTube, as mentioned, custom intent (people who searched certain terms recently) is gold for intent, while In-Market audiences (users actively looking to buy a product in your category) are great for reaching those in research mode. An advertiser found that after getting initial conversion data, moving from custom intent to broader In-Market and Affinity audiences helped scale YouTube spend while maintaining low CAC​. The algorithm can also expand to similar audiences automatically nowadays, but feeding it good signals is crucial. Don’t forget remarketing: One Demand Gen campaign can specifically target your website visitors or past customers (for cross-sell) with tailored content. While remarketing is not “new demand,” it’s still valuable to re-engage warm prospects with different creative (maybe a social-proof laden ad or a limited-time promo), just watch frequency capping to avoid over-showing.
• Measure success through downstream lift, not just last-click. Upper-funnel campaigns often won’t show a strong ROAS if judged solely by direct conversions in Google Ads. Their impact is assistive, they make people search for your brand later or convert via another channel. So, the metrics that matter here are things like view-through conversions, uplift in brand search volume, and overall conversion lift. Google offers Brand Lift and Search Lift studies: you can run a Brand Lift to survey ad-exposed vs. control groups for brand awareness metrics, and a Search Lift to see increased branded searches. Even without formal studies, you can monitor Google Trends or your own Google Ads data: do you see an increase in impressions and clicks on your brand campaigns when YouTube spend goes up? One practitioner recommends looking at lift in brand searches in Google Ads and Search Console as evidence that YouTube brand campaigns are working​. Another “halo effect” to watch is improvements in your other campaigns’ performance, for instance, higher click-through rates on your search ads because more users recognize your brand from seeing a video​. You could even do region-based experiments (run YouTube heavy in one market vs. a holdout market) to quantify the difference. The bottom line: demand gen’s ROI is often indirect. A well-run YouTube campaign might have a 0.5x direct ROAS, but if it led to a 20% lift in organic and brand conversions, the true ROI could be excellent. Use metrics like CPV (cost per view), view-through conversions, and assisted conversion data in Google Analytics to get the full picture. And of course, keep an eye on new customer acquisition from these campaigns, often their biggest value is introducing new people to your brand (feeding the funnel top that later trickles down into the bottom). By allocating 10–20% here, you ensure a steady inflow of future converters, which sustains growth once pure search capture has maxed out.

In essence, the Demand Gen/YouTube component of the structure acknowledges that to scale beyond the confines of search volume, you must invest in creating demand and capturing attention where people spend time (YouTube, Discover, Gmail, etc.). In 2020, many DTC brands ignored these “upper funnel” channels due to attribution challenges. In 2025, savvy marketers embrace them, using better content and better measurement to integrate them into the growth engine.

Finally, remember that these percentages (50–70% Shopping, 10–20% Brand, 10–20% Demand Gen) are guides. A newer brand might spend more on Demand Gen to build awareness, while a well-known brand might put more into Shopping and less into brand (if organic handles it). The core idea is each piece has a role: Capture existing demand (Shopping/PMax), defend and monetize your brand interest (Brand Search), and create new demand (YouTube/Demand Gen). With this structure, you avoid the pitfalls of 2020-style accounts (either too fragmented or too lumped together) and align with how Google’s algorithm works post-2024.

Modern paid search management demands looking beyond surface-level metrics like “Google ROAS” or total conversion count. To truly gauge success and make optimizations that grow the business, savvy marketers in 2025 focus on a few key metrics and KPIs that capture incrementality, customer growth, and market share. Here are the metrics that matter now and how to track them:

• True Incremental Performance (Incremental ROAS or Lift): Instead of obsessing over the ROAS reported in the Google Ads dashboard, the critical question is how much revenue did my ads drive that wouldn’t have happened otherwise? This is incremental ROAS ,  essentially the return on ad spend after accounting for conversions that would happen without ads. As discussed, platform ROAS is inflated by inherent bias​. So, how to measure incremental performance? One way is through Conversion Lift experiments in Google Ads (for larger campaigns, Google can hold out a random control group and measure additional conversions caused by ads). Another is geo experiments (turn off ads in a few regions and compare sales vs. regions with ads on). If those are not feasible continuously, at least periodically conduct tests to calibrate your campaign incremental value. You might find, for example, that your non-brand search campaigns have a 70% incremental factor (30% of conversions would happen anyway via organic or other means), whereas your remarketing display campaign might only be 20% incremental (80% would buy anyway). This helps you allocate budget to high incremental channels. Incremental ROAS can be calculated by multiplying reported ROAS by the incrementality percentage. Some advanced marketers even apply a “credit adjustment” to Google’s numbers. For instance, analytics firm Sellforte suggests using calibration multipliers ,  if you know a channel’s conversions are, say, 50% incremental, you divide Google’s reported ROAS by 2 to get the true ROAS. This way, you’re comparing apples to apples when judging performance across channels. The point is to weave incrementality into your KPIs. If you don’t, you risk scaling campaigns that look good but aren’t growing your business. A practical tip: look at your total company revenue vs. ad spend over time ,  if your Google spend doubled but total sales only rose 10%, that’s a red flag the incremental ROAS is low (the ads are mostly cannibalizing existing sales). On the flip side, if turning off a campaign makes your total sales drop significantly, that campaign had strong incrementality. In 2025, many teams set goals like “Incremental CPA” or use metrics like CAC (customer acquisition cost) blended, which factor in how ads drive overall new revenue, not just last-click. The north star is to maximize net profit and growth, not just make Google’s dashboard look pretty. Keeping a relentless focus on incremental lift steers you toward tactics that truly work (and often saves budget from being wasted on vanity ROAS).
• New Customer Acquisition & Efficiency: Another top metric is new customer volume and cost. Since growth ultimately comes from acquiring new customers, marketers now measure how many first-time buyers each campaign or channel delivers, and at what cost. Google Ads thankfully provides New Customers metrics (if you import your customer list and enable New Customer acquisition tracking, you get columns for “New customer conversions” and even a “New customer conversion value” if you set a bonus value). Keep a close eye on the ratio of new to repeat customers coming from your campaigns. For instance, a brand campaign might only bring 5% new customers (mostly existing returning buyers), whereas a generic non-brand campaign might bring 80% new customers. This drastically changes the value of those campaigns to your business. Many companies now set a target CPA or ROAS specifically for new customers. For example, “We aim to acquire new customers at a $50 CPA on Google.” Or “Our new customer ROAS must be at least 3x to account for future LTV.” You might even decide to run certain campaigns (like Prospecting PMax or broad search) only if they can meet a new customer goal, otherwise pause them. Google’s NCA feature helps here by allowing you to bid differently for new users, as noted. A savvy approach in 2025 is to assign an extra conversion value for new customers ,  say your average first-time customer is worth $100 in immediate sale and an additional $50 in lifetime value, you can tell Google to add +$50 value for new customer conversions. This makes Smart Bidding naturally favor campaigns and bids that bring new folks. Another metric is New Customer ROAS = (Revenue from new customers / Ad spend). This can be low if you include lifetime value, but it’s a gauge of quality. If your overall ROAS is 5x but New Customer ROAS is only 1x, it means most of your spend is going to existing buyers. Sometimes that’s okay (for retention or upsell), but typically you want a healthy mix. Track Cost per Acquisition (CPA) for new customers too. Compare it to your allowable CAC based on LTV. If it costs you $100 to get a new customer who on average is worth $80 gross profit, you’re upside-down ,  time to adjust strategy. By optimizing for new customer efficiency, brands have realized they can significantly boost the incremental growth driven by paid search. As one PPC expert put it, the cost of advertising is high enough that it makes sense to focus on those who haven’t bought from you before​, use cheaper channels like email to re-engage customers who have. So in reporting, include metrics like “# of New Customers from Google Ads this month” and “New Customer CAC” as primary KPIs. These matter more to the C-suite than raw conversion counts. We’ve seen a big mindset shift: in 2020, a Google Ads manager might brag about 100 conversions at $10 CPA, but if 90 of those were repeat buyers, the CFO might not be impressed. In 2025, you’d rather report 50 truly new customers acquired at $20 CPA ,  that is meaningful growth.
• Product-Level Profitability (POAS / SKU ROI): A granular but vital metric set revolves around profit per product. It’s great to drive revenue, but are you selling the right products from a margin perspective? Advertisers are increasingly analyzing performance at the SKU or product-category level, incorporating gross margins into their evaluation. One approach is to calculate POAS (Profit on Ad Spend) instead of ROAS. This means (Revenue - COGS) / Ad Spend. If you can pipe product cost data into your tracking, you can know profit directly. For example, selling a $100 item that costs $50 to produce ,  a ROAS of 2x would actually be break-even in profit terms (because $100 revenue - $50 cost = $50 profit, which equals the $50 ad spend). Tools and platforms now allow sending profit as the conversion value rather than revenue. SavvyRevenue notes that this is a “natural progression” ,  feeding profit into Google’s algorithm lets it optimize bids toward actual business outcomes. Even if you don’t go that far, at least segment your campaign results by product categories with different margin profiles. You might discover that one category with a 5x ROAS is actually less profitable than another with a 3x ROAS, because the latter has double the margins. Historically, some advertisers tried to solve this by segmenting campaigns by margin tiers (e.g. high-margin products in one campaign with lower ROAS target and vice versa). However, one fascinating finding is that customers often buy a different product than they clicked on. Studies show as much as 60% of users buy a different product than the one they initially clicked in a Shopping ad, and up to 40% buy from a different category entirely. This means you can’t just judge a click’s value by that product’s margin ,  cross-sells happen. Therefore, a more holistic view is needed. Many marketers run profitability reports that aggregate total ad cost vs. total gross profit per product or category. If product A shows $500 ad cost and $1000 gross profit from all sales it generated, that’s a 2:1 POAS (maybe not good if target was 3:1). If product B has $300 cost and $900 profit, that’s 3:1 ,  a better performer. These insights guide budget allocation: push the products with the best profit yield, not just revenue. Additionally, watch attach rates and AOV (average order value). If certain products often lead to bigger orders (customers add more to cart), they effectively carry more value. Consider that in your goals. In 2025, Google is rolling out features like Profit-aware bidding in PMax where you can feed COGS into Merchant Center and bid on profit directly​. If you trust Google with that data (and some do, as they believe Google won’t misuse it), it’s worth testing. Otherwise, you can implement your own rules: for example, exclude products from ads that have too low a margin or that consistently spend more than they earn. Some retailers set up an automated rule: “if SKU spent over $X with no sales or with ROAS below Y, reduce its bid or exclude it.” Feed management tools (like we’ll discuss in Section 5) can help make these adjustments en masse. The big picture: Optimize for profit, not just revenue. A campaign hitting a 300% ROAS might sound good until you realize it’s all low-margin items ,  after costs you made nothing. It’s far better to hit a profit goal. If you haven’t yet, start integrating cost of goods and even variable costs (like fulfillment fees) into your Google Ads evaluation. This is advanced stuff that very few were doing in 2020, but by 2025 it’s become a competitive advantage. As one resource stated, tracking profit per transaction allows you to “bid the exact amount you can afford” for each click, instead of relying on blunt average ROAS targets. This granular profit lens will highlight opportunities (and money-pits) that you’d miss if you only look at blended ROAS.
• Search Impression Share & Share of Voice: In a world where capturing existing demand is competitive, Search Impression Share (IS) has emerged as a critical metric. Impression Share is the percentage of times your ad was shown out of the total opportunities it could have been shown for your targeting. In plain terms, it tells you how much of the market’s search volume you’re capturing. A high impression share (close to 90-100%) on your most important keywords means you are dominating those auctions. A low impression share (say 40%) means you’re missing out on more than half the potential customers searching your keywords. There are two main forms: Lost IS (Budget) and Lost IS (Rank). Lost IS (Budget) means your budget is too low to show in all auctions; Lost IS (Rank) means your Ad Rank (a function of bid and quality) was too low to win all auctions. Why does this matter more than ever? Because absolute ROAS alone can be misleading. You might have a campaign hitting a very efficient ROAS, but if its impression share is only 30%, you’re leaving growth on the table. You could likely lower the ROAS target (i.e. bid more) or increase budget and capture more volume until that ROAS starts to drop to an acceptable level. Many brands have realized that once you have a profitable model, you should scale until you saturate the market ,  maximize impression share until the marginal return dips. For instance, if your non-brand search campaign has 50% impression share and is exceeding your ROAS goal, consider loosening that goal or adding budget. As Adalysis notes, if an automated campaign is working well and still losing impression share due to budget, you generally want to raise your budgets (or lower your ROAS targets) to capture the additional volume. Conversely, if you have high lost IS due to rank, you might need to improve quality or increase bids (which could reduce ROAS) to appear more often. The strategic view is to decide where to trade off ROAS for volume. Absolute ROAS as a metric doesn’t reflect opportunity size. Impression Share does. For example, let’s say Campaign A has ROAS 5.0 and 50% IS, Campaign B has ROAS 3.0 and 95% IS. Campaign A looks “better” by ROAS, but it’s only reaching half the market. Campaign B is milking its market fully. It might make sense to actually invest more in Campaign A until IS increases and ROAS falls closer to, say, 3.0 ,  because then you’re getting more total profit even though efficiency per conversion is a bit lower. In 2025, many marketers focus on Incremental profit: would an extra $1 in spend produce more than $1 in profit? If yes, keep spending until it doesn’t. Impression share is a guide here ,  if you’re below 100% and each incremental chunk of spend is still profitable, go for it until that stops. This is how you push past growth plateaus. It’s also a defensive metric: if your impression share on core terms slips, likely competitors are stealing share. For brand campaigns, you want ~90%+ IS (some loss due to rank is fine if you’re saving money on irrelevant long-tails, but monitor it). For important generic terms, track IS weekly. It might vary with seasonality and budget changes. A big increase in Lost IS (Rank) could mean a new competitor entered or someone raised bids ,  a signal to revisit your strategy on those terms. Market share in the SERPs is as much a goal now as efficiency. Especially when you have fixed budgets, you must allocate such that you maximize impression share where it counts most. As an example, if you’re hitting your daily budget and seeing 30% Lost IS (Budget), that’s usually an argument to raise the budget or shift spend from elsewhere, assuming CPAs are good. Internally, you can set targets like “Maintain >80% impression share on all top 10 non-brand keywords”. This ensures you aren’t inadvertently ceding ground to competitors just to hit a ROAS number. Ultimately, higher impression share (at a sustainable ROAS) means more revenue and more market dominance. It’s a metric that marries well with incremental thinking ,  it pushes you to capture all the truly incremental conversions available in your category.

In short, 2025’s performance marketers care about incremental growth, new customer counts, profit, and market share ,  not just the vanilla metrics of yesteryear. By focusing on Incremental ROAS vs. reported, New vs. Repeat customer metrics, SKU-level profit, and Impression Share, you gain a 360° view of program health. These metrics guide better decisions: e.g. to cut a campaign that’s mostly cannibalizing, or to double down on a category that’s printing profit, or to move budget to capture missed opportunities. As you adopt these metrics, educate your team and leadership about them. It represents a maturing of digital marketing: we’re aligning ad success with business success more than ever. The days of blindly optimizing to in-platform metrics are gone ,  we optimize to real business outcomes now.

Even with automation doing heavy lifting in bidding and targeting, human oversight and optimization are still critical. A senior performance marketer should have a weekly ritual of checks and tweaks to keep the account in top shape. Think of it as maintenance that ensures the algorithm’s “training data” and user experience remain optimal. Here are the weekly optimization tasks you should be doing, and some advanced tips for each:

• Landing Page Performance Audits: Each week, review how your landing pages are performing, because they directly affect conversion rates and Quality Score. Use Google Analytics or GA4 to check bounce rates and conversion rates for traffic from your Google Ads campaigns. Identify any pages where users drop off or convert poorly. For example, if your Shopping ads all land on product pages, see which product pages have high bounce or low time-on-site,perhaps the price is too high or the content is lacking. Consider running A/B tests on important landing pages (using tools like Google Optimize,although Optimize was sunset, you can use GA4’s new experimentation or third-party tools). Test things like simplifying the page, adding trust badges, changing headlines to match the ad query (message match). Google Ads also provides a “Landing page experience” rating in Quality Score,if you see keywords with low landing page experience, that’s a red flag. Ensure your pages load fast (PageSpeed Insights should be in the green; slow pages kill conversion and mobile experience). Ensure they are mobile-friendly (check on your own phone). If a campaign has multiple ads going to different pages (e.g. different product categories), compare their performance and allocate traffic to the better ones. Pro tip: incorporate relevant keywords on the landing page (organically) that align with the ad. This can boost Quality Score and relevance. For instance, if you have an ad group for “mens running shoes” but your landing page title says “Acme Sports - Shoes”, consider changing it to “Men’s Running Shoes - Acme Sports” so the user immediately sees it matches their search. Each week, prioritize a couple of pages to tweak and measure results next week. Over time these incremental improvements add up to significant gains in conversion rate (even a 0.5% absolute increase in CVR can allow you to bid more or get more ROI). Never adopt a “set and forget” mindset for landing pages; user behavior and expectations change, and a page that worked last year might need refreshing now.
• Sitelink & Extension Effectiveness: Your ad extensions (sitelinks, callouts, structured snippets, etc.) are important components that can influence CTR and conversion path. Every week, look at your Ads & Extensions report in Google Ads. For sitelinks, check their individual click-through rates and conversion metrics (Google Ads lets you see how often each sitelink was clicked). Identify which sitelinks are seldom clicked,maybe they aren’t attractive or relevant. Try updating their text to be more compelling (e.g. add a value prop like “20% Off New Arrivals” instead of a generic “New Arrivals”). Also ensure the sitelink landing pages are working and relevant. Sitelinks not only drive extra clicks, they also make your ad larger,which tends to improve overall CTR. Google’s research shows using ad extensions can increase CTR by 10–15% on average. So, you want to maximize that advantage. Each week, you might test a new sitelink or swap one out for a seasonal offer. Similarly, review your callout extensions (those short phrases),are they up-to-date? If you have “Free Shipping over $50” as a callout but recently changed it to free shipping on all orders, update the extension. Structured snippets (like “Brands: Nike, Adidas, Puma…”) should reflect current inventory or offerings. Also pay attention to extensions at the campaign vs. account level. Maybe your account-level sitelinks are generic, but a specific campaign could use more specific ones. For example, a “Holiday Sale” campaign should have sitelinks for “Holiday Gifts” or “Black Friday Deals”. Implement that and pause the generic ones for that campaign. Over time, extension performance can decline due to banner blindness,so refresh them periodically (new wording, new offers). Remember, extensions contribute to Quality Score’s expected CTR component indirectly; an ad with full extensions is more likely to get a higher CTR, which Google rewards. If you notice an extension with low CTR that isn’t necessary, consider removing it,sometimes too many extensions can dilute the message. However, generally more is better, as long as relevant. So weekly, it’s about optimization and relevance: keep extensions fresh, prune underperformers, and add any new ones that make sense (Google keeps adding formats,e.g. image extensions, promotion extensions, etc.,use them if applicable). They cost nothing extra per click but can improve results significantly.
• Asset Performance Review (PMax and RSA): If you’re running Performance Max or Responsive Search Ads in search, Google provides asset performance ratings that you should monitor. In PMax asset groups, for each text, image, and video asset you’ve provided, Google will label them “Low”, “Good”, or “Best” based on their relative performance. Each week (or bi-weekly if volume is low), check the PMax asset report. Any asset marked “Low” is underperforming compared to others. Plan to replace those with new variants. For instance, if a headline “Shop Quality Widgets Online” is rated Low, try a different angle like “Premium Widgets,Free Shipping” or include a promo. If an image is Low, perhaps the creative didn’t resonate,try a different product shot or a lifestyle image. Over a few weeks, you’ll refine your asset mix such that most are rated “Good” or “Best”. Google recommends regularly replacing Low assets to keep improving ad strength. The same concept applies to Responsive Search Ads (RSAs) in search campaigns: check the Ad Strength score and asset combination performance. Ad Strength is a measure of how well your RSA follows best practices (variety of text, relevant keywords, etc.), rated from Poor to Excellent. Aim for “Excellent” ad strength on all RSAs. Google’s data shows that moving an RSA from “Poor” to “Excellent” ad strength can drive 12% more conversions. That’s a free uplift for just tweaking ad copy! So each week, pick a couple of RSAs with less than Excellent strength, and implement the suggestions Google gives (e.g. “Add more headlines,” “Make headlines more unique”). Also look at which combinations of headlines/descriptions are actually showing (Google Ads has an asset details view that shows top performing combos). If a certain headline isn’t showing much or when it does, performance lags, consider rephrasing it or replacing it. Feed the machine new creative: The algorithm can get “creative fatigue” if the same assets run too long, especially on display/video. So swapping new images or refreshing text can prevent ad fatigue in the audience as well. If you have video assets, see their view rates or performance,a Low rated video might indicate it’s not engaging; maybe edit it or test a new video. The goal is a constant creative optimization loop: use Google’s asset feedback to evolve your messaging. It’s almost like multivariate testing at scale,Google tells you what works; your job is to double down on what works and fix or drop what doesn’t. Keep a log of changes and watch how performance shifts week to week as you optimize assets.
• Product Feed and PLAs Health Check: If you run Shopping or PMax, your product feed is the lifeblood. Each week, log in to Google Merchant Center and check the Diagnostics tab. Look for any disapproved or restricted products, and fix those immediately. Common issues could be price mismatches (your website shows a different price than feed,update the feed), policy violations (maybe a description has forbidden words), or GTIN errors. Also check for expiring feeds or feed fetch errors. If your feed didn’t update, you could be showing out-of-stock items, which is wasted spend and bad user experience. Next, review performance by product or product group in Google Ads. Identify products that spent a lot but haven’t sold in the last week,consider lowering their bid or excluding them if this trend continues (could be they need better images or pricing). Conversely, see if any product is consistently selling out or low on stock,you might downweight it in ads to avoid overselling (nothing worse than paying for clicks to a product that’s back-ordered). Check your search terms report for Shopping ads (now available via Reports in Google Ads),see what queries trigger your products and if there are irrelevant ones. Add negative keywords at the campaign level for any consistently irrelevant queries that slip in (though Shopping uses the feed, Google will still match broadly sometimes). For example, if you sell luxury leather bags and see queries for “cheap knockoff bags” leading to clicks, you might negative “cheap” or “knockoff”. Also, evaluate the Product Titles and Images as they appear on Shopping. One trick: search some of your keywords in an incognito window and see how your product ads look among competitors. Is your title truncated or less descriptive? Is your image dull compared to others? These insights can drive feed improvements,maybe shorten titles or put key info first (since often only the first 30-70 characters show). Over the week, also check any Promotions or Sale prices in the feed,if a sale ended, remove the sale price in the feed to avoid mis-pricing. If you have custom labels (e.g. “clearance” items), ensure you updated them for items that changed status. In summary, a weekly feed review catches any data issues early and keeps your Shopping ads running smoothly. Many brands using feed automation set it and forget it, but winners monitor it like a hawk,ensuring feed quality, accuracy, and relevance continuously. It’s not glamorous, but it can prevent huge waste (imagine a feed issue causes half your products to be disapproved,if you catch it in a week vs. a month, you save potentially thousands of lost impressions).
• Bid and Budget Adjustments & Anomaly Detection: While much bidding is automated now, you still should review how the smart bidding strategies are performing and tweak campaign settings or budgets. Each week, scan campaign-level performance. Did any campaign suddenly spend way more or less than expected? Investigate why. If ROAS or CPA skews heavily for a week, check if there were external factors (holiday sale? tracking issues? site outage?). If a Target ROAS campaign keeps missing its target, perhaps the target is set too high,consider easing it a bit to let it capture more conversions. On budgets: see if any campaigns are hitting budget caps consistently (limited by budget),if they are profitable, you likely should raise the budget to let them breathe (don’t constrain a winning campaign). Conversely, if a campaign underspent far below budget, maybe search volume was low or the target/bid too restrictive,check if something’s off (did an important keyword get paused accidentally?). Watch Impression Share changes week over week as mentioned. If you see Lost IS (Budget) high in a profitable campaign, increase budget. If Lost IS (Rank) high and you can afford a lower ROAS, consider raising bids or target (or improving quality via better ads). Also, look at device and audience performance if using tROAS/tCPA,Google auto-adjusts, but you might catch insights (e.g. mobile converting much worse,is the mobile site okay? Fix that rather than cut mobile bids, since smart bidding might not fully account for site experience issues). Enable the creative learning and bid strategy learning insights: Google Ads will often flag if, say, “Your ROAS target might be limiting conversions” in the Recommendations tab. Evaluate those suggestions critically but they can indicate where the algorithm is struggling. For anomaly detection, consider using automated rules or scripts that alert you if, for example, spend or CPA deviates by >30% from the prior week’s average,that way you can quickly react instead of finding out days later. Weekly meetings should cover these anomalies: e.g. “Search Campaign X’s CPL spiked 40%,we found it was due to a competitor dropping prices (conversion rate fell); we adjusted our ad copy to emphasize price match”. This kind of vigilance keeps performance on track. The algorithms are good, but they’re not omniscient,a human needs to diagnose why something changed and feed those insights back in (maybe through new negatives, bid cap, budget shifts, etc.).

By conducting these weekly optimization steps,landing page tweaks, extension updates, asset refreshes, feed management, and bid/budget oversight, you create a feedback loop that enhances the effectiveness of Google’s automation. Think of it as grooming the system: you’re removing friction points and giving the machine the best conditions to succeed. Many brands that struggled with Google’s automation found that a lack of these human touchpoints caused issues to compound (e.g. one unchecked broken page can tank Smart Bidding performance). With a disciplined weekly checklist, you’ll catch problems early and continuously improve results. This also gives you actionable insights to report: instead of just “we hit our numbers,” you can say “we improved CTR 5% by testing new sitelinks” or “we increased conversion rate on our top page after fixing mobile load speed,resulting in 10% more conversions this week.” These tangible improvements are the hallmark of a well-run account in 2025.

In e-commerce especially, product feed management is the secret sauce behind high-performing Google campaigns. Your Merchant Center feed doesn’t just list products, it’s a rich dataset that Google’s algorithms use to match your ads to searches and to determine relevance. Thus, optimizing the feed can yield significant gains in impressions and conversion efficiency. By 2025, feed management has grown into a sophisticated practice with specialized tools. Let’s explore the key strategies for feed excellence:

• Automate and enrich feeds with tools: Manually managing a feed (especially with hundreds or thousands of SKUs) is error-prone and time-consuming. Feed management platforms like GoDataFeed have become invaluable. These tools connect to your e-commerce platform (Shopify, Magento, etc.), pull your product data, and let you transform and optimize it before it reaches Merchant Center. For example, you can use GoDataFeed to append keywords to titles, generate custom labels, or exclude certain products based on rules. They also ensure your feed updates regularly so inventory and prices stay in sync. According to GoDataFeed, they “automate feed updates, optimize product data for Google’s algorithm, and ensure visibility across Google’s shopping ecosystem.”​ In practice, this means if you want to add “Free Shipping” to all titles of products over $50, you can do it in one rule rather than editing each title. Or if your brand name is missing from titles, you can prepend it (e.g. “Acme Widget, Blue” instead of just “Blue Widget”). These optimizations directly affect how often and where your products show up. Another tool mention is Marpipe, which focuses on creative automation (generating many ad creative variations). While Marpipe is more for display and social ad creatives, they are expanding into dynamic product ad creative. Marpipe uses automation and AI to generate multitudes of ad variants and test them​. It can create every permutation of an ad with different backgrounds, text, etc. While not directly a feed tool, it can integrate with product catalogs to produce dynamic creatives (think dynamic remarketing ads with creative templates). Using these tools, a small team can achieve what used to require a dedicated IT or design team. The point is harness automation to keep your feed optimized at scale. If you have a frequently changing catalog, a tool like this is almost mandatory to avoid feed downtime or stale data. And with the time saved, you can focus on strategic enhancements (like experimenting with title formulas or custom labels) rather than fighting fires with feed errors.
• Optimize product images for advertising: By default, your product images might be simply pulled from your website. However, the images that work best on a website are not always the most compelling in an ad context. On your site, a plain white-background image might suffice. But in a competitive Shopping ad carousel, an image that stands out can drive a higher CTR. Consider enhancing images specifically for Merchant Center: ensure high resolution (at least 1000x1000 for apparel, etc.), clear lighting, and product fills as much of frame as possible. Avoid tiny product in a big white box, users scroll past those. If your platform (like basic Shopify) only has one image per product, think of adding more variant images to the feed (Google allows additional_images for Shopping that users can click through). For example, include an image of the product in use, or from another angle. However, be mindful of Google’s policies: no promotional text or watermarks on the main image, and usually a plain background is recommended for the primary image. But you can get creative with secondary images. Some merchants create lifestyle imagery specifically for use in Performance Max or Discovery ads, showing the product in context, which can then be fetched via the feed for those placements. If your brand is aesthetic (fashion, home decor), these visuals are crucial. Also, in 2025 Google allows image extensions in search ads and uses your Merchant Center images in various formats. So having a rich set of images improves your chances of showing in visual ad formats. Beyond Shopify’s standard setup implies doing more than the bare minimum of one image and a title. It might mean customizing thumbnails (e.g. crop to show details), or ensuring consistency (all images same aspect ratio to not get downsized weirdly). Some advanced retailers even A/B test their product images by swapping them in the feed and seeing impact on CTR. For example, one might test a product photo on a plain background vs. one on a subtle textured background to see which attracts more clicks (Google Shopping will reflect CTR differences). Use feed rules or content APIs to swap images for testing. Remember: the image is often the first (and sometimes only) thing users judge in Shopping ads. Make it count.
• Provide complete and correct GTIN data: We touched on GTIN (Global Trade Item Number) earlier in metrics. Including the correct GTIN for each product (the barcode number from the manufacturer) is extremely important in Google’s eyes. Google uses GTINs to group identical products across sellers. If you provide the GTIN, your product ad becomes eligible to show in product listing pages and comparisons, and Google’s algorithm gets a confidence boost in your product data. Google explicitly flags products without GTIN (if they are known to have one) as “Limited performance due to missing identifiers”. Having GTIN can increase impressions significantly. In fact, having valid GTINs helps in queries like “best [product]” because Google can aggregate reviews for that GTIN and show star ratings. DataFeedWatch noted that “GTINs also get you placement within searches like ‘Best’ and ‘Top’” and without them you may not appear in those high-intent searches​. If your products are custom or don’t have GTINs (handmade, vintage), you can mark them as such (identifier_exists = FALSE) to avoid penalties. But for anything mass-manufactured, hunt down those codes. If you sell branded items, coordinate with suppliers to get GTINs for all. The payoff is better matching, when a user searches a specific model or UPC, your product will match exactly if GTIN is present. Additionally, GTIN helps Google merge reviews, the Product Ratings program uses GTIN to aggregate reviews across sellers. So a product with GTIN can show star ratings (if the aggregate hits the threshold), which boosts CTR. It also helps prevent disapprovals due to mismatched identifiers. This is a one-time setup thing (get all your GTINs in the feed), but review weekly if any new products lack it. If you added new SKUs and forgot GTIN, update them. Also ensure brand and MPN are filled where relevant. These identifiers together increase Google’s “confidence” in your product data, which often translates to better ad placements. In short, don’t skip the catalog housekeeping, complete GTIN data is a must for superior Shopping performance​.
• Manage and leverage product reviews and ratings: Social proof can make or break an ad’s effectiveness. Google Shopping allows star ratings on product ads (distinct from seller ratings which are for the store). These product ratings come from aggregated customer reviews. To get them, you need to either collect reviews on your own site and submit to Google via feed, or use an approved third-party review aggregator (like Yotpo, Trustpilot, etc.) that shares reviews with Google. Actively managing reviews means: encourage your customers to leave reviews post-purchase, address negative feedback (customer service follow-up, etc.), and ensure those reviews get to Google. The benefit is huge: ads with star ratings grab attention and instill trust. An Omnitail study cited by DataFeedWatch found a 107% increase in CTR for ads with seller ratings vs. without. That’s more than double the clicks! Similarly, product-level ratings likely have a significant uplift (Google hasn’t published a specific stat for that, but it stands to reason they help a lot). So if your products have good reviews, make sure they’re visible. To do this, maintain a feed of product reviews to Google Merchant Center or use a facilitator. Google requires at least 50 reviews across your products to start showing stars, and each product generally needs 3+ reviews. Make it a goal to get every product past that threshold. If some products have 1-2 stars (poor ratings), that’s tricky, you might actually opt not to show those, but Google doesn’t let you selectively hide product ratings. Instead, work to improve those products or their descriptions to set better expectations. On the flipside, if you have many happy customers, consider enabling the Google Customer Reviews program (free) to collect verified reviews in exchange for a badge. Also use your reviews in your ads: e.g. in Marketing copy (like callout extension “500+ 5-star reviews from customers”). Every week, glance at new reviews coming in. Respond to issues to potentially turn a 2-star review into a 4-star. Also, update your product feeds if a product’s rating average changes (though Google mostly handles this). Think of reviews as an extension of feed optimization, it’s user-generated content that can significantly enhance your ad’s appeal. The ROI is clear: more trust = more clicks = more conversions, as long as the reviews are positive.
• Promotions and offers via Merchant Center: Running special promotions can set you apart in Shopping results. Google Merchant Center allows you to submit promotions (like a coupon code, discount sale, free gift, etc.) which then show up as an “offer tag” on your product listing (a little highlight like “15% off code ABC” or a strikethrough price). Weekly, you should plan and update any promotions you want to feature. For example, if you have a Spring Sale 20% off, submit that promotion in GMC (including promo code and dates). Google will then show a clipped coupon icon or sale annotation on your ads. This can boost CTR and conversion rate significantly by conveying a deal. A Google study found that promotions led to a 28% lift in conversion rates during a sale. That’s huge. So use this feature. It’s underutilized by many advertisers (often only big retailers do, but smaller ones can too). You can schedule promos ahead of time. Also consider using the sale_price attribute in your feed when an item is on sale, Google will show the original price crossed out and the sale price, another eye-catcher. Just ensure to remove it when the sale ends to avoid “price mismatch” issues. If you run promotions frequently, consider setting up a Promotions Feed (a spreadsheet of all promos)​ for easier management. Each week, verify that any ongoing promo is accurately reflected (dates correct, code working). If reviews are trust, promotions are urgency/appeal, both are important. Combining both (a well-rated product on sale) is ideal. Additionally, use the Merchant Center Promotions Dashboard to see performance metrics of your promos, did the CTR improve? How many redemptions? This can inform which types of offers resonate (free shipping vs. 10% off, etc.).
• Routine profitability and performance audits: Just as you optimize campaigns, you should audit your product catalog’s performance regularly. On a weekly basis, identify which products or categories are unprofitable with current ad spend (as touched on earlier). You might create a custom report: list of products, their ad spend, and their sales (conversion value) for the week or month. Flag those with Ad Spend > Gross Profit. For any losers, decide an action: reduce bids (maybe exclude from primary campaigns and advertise only via a separate low-bid campaign), improve their conversion rate (maybe the product page needs better info or reviews), or accept them as loss leaders if they have LTV upside (e.g. a product might be a gateway to new customers who later buy more). Also flag big winners: products with high ROI, can you scale them further (stock permitting)? Feed management excellence is not just avoiding errors, but actively steering spend to the best opportunities. Some merchants set up priority campaigns for their top sellers to ensure those always get full coverage. Your feed can include a custom label like “bestseller” and then a campaign exclusively for those with higher budget/tROAS flexibility. Meanwhile, “low performers” might be in another campaign with conservative settings. This segmentation is done via feed labels, which you adjust as performance shifts. So yes, feeds aren’t static, you might recategorize products based on performance signals. For instance, tag items that had <1 ROAS in last 30 days as “Low_ROAS” via a supplemental feed, then use a rule to exclude or lower their priority. These kinds of dynamic feed adjustments (some do it daily, but weekly could suffice) keep your ads efficient. It’s like pruning a garden: cut the weeds (money-wasting products) and water the flowers (profitable products). A weekly audit might also involve checking competitor pricing, if you have a price competitiveness data (Google’s auction insights or the Shopping tab), and you see you’re way above market price on an item, you either adjust price or realize that product may not sell via ads (people will choose the cheaper competitor). Feed can include a sale or a special highlight if you match a competitor’s price, etc.

In summary, feed management excellence is about merging technical accuracy with strategic enhancement. By automating the grunt work, polishing your product data (titles, images, GTIN, etc.), harnessing social proof and promotions, and continuously aligning the feed with performance goals, you turn your feed into a competitive advantage. In 2020, many advertisers just uploaded a basic feed and left it. In 2025, the feed is a living, optimized asset. It’s worth investing time and resources here because every improvement can boost your Shopping and PMax results across thousands of auctions. A well-managed feed means your ads show up in the right searches, look enticing, and drive profitable clicks, which is exactly what we want.