Most companies in these interviews run paid media inside the ad platforms themselves or through an agency, and the platforms are the tools they rate. Google Ads sits above what buyers give software generally; Meta Ads, LinkedIn Ads and the programmatic platforms are not far behind. The regrets are not about which platform to use. They are about running them: how much the automation shows, whose numbers to believe, what the auction costs, and who answers when something breaks. For how companies organise the work and when they add a tool on top, see what paid media management software is; for the creative layer, what creative analytics software is.
What do paid media buyers wish they’d known?
Seven things recur. Automated campaigns such as Performance Max and Advantage+ save time and often perform, but hide what is working, which is hard when you have to explain the results. The platforms’ recommendations and default settings are not always in the advertiser’s interest, so buyers test them. Each platform assigns its own conversion credit, which is not the same as incremental sales, and view-through conversions are where its numbers and yours most often part. Buyers rate the tool and the auction separately: a platform can score a nine while its costs rise. Support from the platforms is uneven and depends on the account. The algorithm depends on the conversion signals it receives and models. And a programmatic platform needs an owner, time to optimise and a clear view of fees.
The seven lessons
Automation buys efficiency with your visibility
Google’s Performance Max and Meta’s Advantage+ hand targeting, bidding and placement to the platform. Buyers describe the trade from both sides. A large telecommunications company says Advantage+ gave its team “more time on our team’s hands, less manual labor, increased efficiencies,” and finds “There’s really no real drawback of something like Advantage Plus, except that it is quite a black box for us.” An insurance company that rates Google a ten says “we are very comfortable with the black box,” “as long as you’re producing results on the other side.” A luxury brand’s media director goes further: “The algorithm is consistently outperforming any data signals or targeting I use around my CRM data.”
The cost lands on whoever has to explain the results. A consumer-finance company says of Performance Max, “it still seems like a black box. We can’t see where ads are being served.” A national retailer’s planning director describes the push toward “more algorithmic-driven things” as making it “challenging when you’re the person responsible for explaining the performance.” A DTC fashion brand says Google has “become incredibly black box” and that its team now needs “a lot of workarounds that are not Google recommended” to keep a distinct product strategy. A European bank sums it up: “It is a black box. You don’t really know what is going on behind, but you don’t have much of a choice.” Google’s help centre now lists a channel performance report, search terms and placement reports, campaign-level negative keywords and brand exclusions for Performance Max. They help investigate and restrict where ads run, within limits Google states: budget cannot be set per channel, and placement reports are not for judging performance. The buyers above, all interviewed in 2026, describe the gap that remains, not an absence of controls.
Named in this context: Performance Max · Advantage+ · Google Ads · Meta Ads
Test the platform’s recommendations before you accept them
Two buyers describe the platform’s own suggestions making results worse. The national retailer above says of Google Ads that “you’ll even see the results get worse when you enable some of these recommendations that they tell you to do within the platform.” A media agency that rates Meta Ads a three describes features that switch on when a campaign launches: “a lot of times we’ve experimented and seen that the campaigns run better without them.” Among them is an automated creative option that, in its account, “just creates creatives on its own without taking any sort of recommendation or approval from the user,” leaving ads with errors nobody approved.
Those are two companies’ results, not a measured effect. The implication — ours, not a buyer’s — is to review recommendations and automated settings before accepting them, test material changes against a comparison where you can, and record what changed in creative, targeting, placements and measurement. Some options switch on by default, and what is available varies by campaign type and account, so check your own setup. Neither a platform’s recommendation nor one advertiser’s bad experience shows that a setting helps or hurts every campaign.
Named in this context: Meta Ads · Google Ads · Advantage+
The platform assigns its own conversion credit, which is not incremental sales
Platforms report the conversions they credit to their own campaigns, under their own attribution rules, lookback windows and available signals. That is a different measure from the additional sales the advertising caused, which incrementality tests and mix modelling estimate. The buyers here who checked the platform’s figure against an independent read describe a gap, widest in lower-funnel campaigns and view-through credit. An apparel company ran incrementality tests with a third-party platform and found “the ROAS and attributed revenue from the platform are largely inflated” for Google, and that its lower-funnel search and Performance Max spend is “not nearly as effective as the platform reports.” On The Trade Desk its display results came out so far below the platform’s own figures that it has been pulling back display budget, and the culprit it names is “a lot of conversions attributed to view-through attribution, which we don’t have a lot of confidence in.” A wellness-products brand says “we know that Google is over-attributing,” even when read through its attribution tool.
The view-through gap shows up on Meta too. An online marketplace rates Meta a three because “the issue with Meta is the attribution and the measurement”: its view-through conversions do not reconcile with Google Analytics. It is not cutting Meta, only holding spend where it is. A disagreement like that is not by itself proof of over-attribution: the two systems apply different rules, windows and signals, and neither alone establishes incremental sales. These are specific advertisers’ findings, not evidence that every platform overstates. For how buyers run tests that answer the causal question, see what incrementality testing is.
Named in this context: Google Ads · The Trade Desk · Meta Ads · Performance Max
The rating is for the tool; the cost is the auction
Buyers keep the two apart. A retail bank rates Google Ads a nine and adds its one complaint: “the marketing CPCs are getting higher on a day-to-day basis. But as a tool, it’s like nine.” A wellness-products brand names “The rising costs of all platforms” as its biggest acquisition challenge, because “all these CPM costs are rising,” and says Google is getting more expensive than Meta for non-branded search: “You need to have a good AOV for Google ads to make sense for non-branded campaigns.” A moving-and-storage company’s 2026 forecast has its paid-search acquisition cost rising by about a fifth, and an online-learning company says buying on Google and Meta is “becoming more expensive because of the AI.”
LinkedIn draws the clearest version. A large software company says “LinkedIn Ads make it easy to target,” and also that “LinkedIn Ads are incredibly expensive and sometimes don’t have the ROI that you want.” Another software company says its “high CPCs and CPMs” make it “difficult to scale with strict acquisition targets.” So a platform’s rating says little about whether its auction suits your margins; that is a question about your order value and your conversion rate, not about the tool.
Named in this context: Google Ads · LinkedIn Ads · Meta Ads
Support is uneven, and the account team is part of what you buy
Where buyers praise a platform’s support, it is a named team. A sporting-goods retailer says its high rating for Google “is based on the team, the support team that we get,” which helps it navigate new betas. The insurance company above calls its Google account service “second to none”, and a digital publisher says “We’re able to meet with our Google reps and remove any barriers.” A pet-store chain reaches Google through its agency: “Our agency is a premier agency, so we really have access to Google through our agency.”
Without that, the experience is different. A small founder-led pet-services company calls Google’s customer service “a terrible, terrible process.” A community bank rates Google’s ads a nine but says “hearing back from somebody at Google takes a long time.” Size is not a guarantee: a large online retailer says of both Google and Meta, “the support when there’s something that goes wrong is not great,” and a consumer health app says “we don’t have any support from Meta to grow.” Newer channels court the account: a wellness brand just starting on TikTok gets “direct support with actual TikTok reps,” and a payments company finds TikTok’s support “quite good relative to our other platform reps.” A home-services company meets its Microsoft Advertising reps regularly and says the team “takes time to look for new products for us.” Our recommendation, not a practice buyers describe: ask who your contact will be, and what spend or agency status it depends on, before you commit budget to a new platform.
Named in this context: Google Ads · Meta Ads · TikTok Ads · Microsoft Advertising
The algorithm depends on the conversion signals it receives and models
Automated bidding works from the conversion events, values and other signals available to it, and Google’s help centre says its consent mode can feed modelled conversions into bidding where visitors decline cookies. Missing or inaccurate signals can weaken optimisation, though a pixel that does not fire does not always leave a platform with no conversion information. A jewellery retailer describes its own setup when visitors decline cookies: “you can’t optimize your paid media ads around Facebook and Google because your ad platforms can’t see conversions fire in the conversion pixel.” Some companies use server-side integrations to improve how events reach the platforms, within their consent and privacy obligations. A DTC furniture brand uses Elevar “for the server-side purchase events to ad platforms,” and a sportswear retailer uses it because “It allows us to run multiple Meta pixels in the back end of our Shopify stores.” Elevar routes conversion data; it is not an attribution tool. And a global beauty group still finds technical problems in both Meta’s and TikTok’s ad managers, “like the pixel not working.”
Learning also takes time. A consumer subscription company that started on The Trade Desk in 2025 says “it always takes time to get it working”: “we came in really hot in terms of a cost per acquisition perspective,” before results improved as it fed the platform its own customer data and better creative. The implication, ours: give a new campaign or platform enough time and conversion volume to judge, without assuming the learning period guarantees improvement, and check the conversion feed first when results drop.
Named in this context: Elevar · Meta Ads · Google Ads · TikTok Ads · The Trade Desk
A DSP needs an owner, time to optimise and a clear view of fees
Programmatic buying platforms such as The Trade Desk and Display & Video 360 buy inventory across exchanges and publishers; they are a different product from the ad consoles and from management tools. Buyers who rate them well still describe the work. A home-improvement company rates The Trade Desk a six: “it is complicated. You need someone owning it. It’s not just like set it and forget it.” An enterprise software company says of Display & Video 360, “If I bring someone on to help manage it, it’s quite a bit of training.” Several buyers here run it through an agency instead.
The second concern is what the advertiser pays for technology and management on top of the media. Depending on the arrangement, that can include the DSP’s own fees, a reseller’s or other intermediary’s charge and the agency’s management fee; buyers sometimes call all of it margin, but they are different charges to compare with the media spend. A consumer-electronics maker says that for Display & Video 360, “not every company gets access; you have to pay a lot of fees. So we have to work with a third-party company,” which takes a percentage of its spend. A distributor moved its programmatic buying from Display & Video 360 to StackAdapt when it changed digital agency, after the old arrangement “was getting quite costly, so the margins were getting quite high”; that move is complete. A restaurant chain calls The Trade Desk’s fees “extremely high compared to other brands, other partners,” while a supermarket chain chose Display & Video 360 and finds The Trade Desk “a much more expensive platform to use, although it does provide a little bit more capability.” Others are happy hands-on: a furniture retailer finds The Trade Desk “really easy to get in there and break things down at the line item level.” And one platform rarely covers everything; an insurer says Amazon’s inventory means “you’re forced to use more than one DSP.”
Named in this context: The Trade Desk · Display & Video 360 · StackAdapt · Amazon
How buyers rate the platforms
Every figure is the average across verified buyer interviews, on published transcripts. The consoles and the programmatic platforms sit close together; the spread in how buyers experience any one of them, from a three to a ten, is wider than the spread between them.
| Product | Rating | What buyers describe |
|---|---|---|
| 8.1 | A creative-analytics tool, not an ad platform. Bought mostly by DTC growth and creative teams to read which ads work; covered on what creative analytics software is. | |
| 7.8 | The most-rated ad platform here. Praised for intent and results; the complaints are automation that hides the detail, rising costs and uneven support. | |
| 7.8 | Google’s programmatic buying platform. Valued for breadth of channels and audiences; described as needing training, and sometimes reached through a reseller. | |
| 7.5 | A programmatic buying platform, often run by an agency. Praised for control at the line-item level; complaints are complexity, fees and view-through credit. | |
| 7.4 | Reach and audience building; Advantage+ splits buyers. Low raters cite settings switched on by default, reporting they cannot reconcile with their analytics, and support when things break. | |
| 7.2 | Precise professional targeting for B2B; the recurring complaint is the cost per click and per impression. | |
| around 7.5 | Too few buyers for a decimal. Bought to reach new and younger audiences; newer advertisers praise direct access to TikTok reps. | |
| around 7.5 | Too few buyers for a decimal. A tool for building paid social across platforms; covered on what paid media management software is. | |
| around 6.5 | Too few buyers for a decimal. Run beside Google for search; buyers say it trails Google’s features and flags policy issues more often. |
Two patterns sit behind the numbers. The programmatic platforms rate close to the consoles, but their raters are more often describing a platform an agency runs for them. And Microsoft Advertising’s lower figure comes with a consistent explanation: a digital publisher finds its system misflags policy issues more often than Google’s and needs several rounds with its reps to clear them, and a home-services company says “the Microsoft Ads platform always runs a little bit behind the Google Ads platform.” Motion and Smartly are not ad platforms but tools run on top of them, bought for one job each; Search Ads 360, Skai and the other management layers are rated by too few buyers here, and their story is on the definition page.
Want this checked against your own stack?
Start my checkup →The stories behind the lessons
The counter-current: buyers complain about the platforms and keep spending on them
For all the complaints above, few buyers here describe leaving Google or Meta. The online marketplace that rates Meta a three is holding its spend, the European bank says it has little choice, and the apparel company that found Google’s figures inflated is considering moving budget between Google’s campaign types rather than off Google. The platforms rate well because they deliver, and the frustrations are about control, measurement and service around that. Our reading, not a buyer’s: the decision most buyers face is how to run the platforms with more of their own measurement and testing, not which platform to replace.
What this means for your paid media
Our recommendations, not practices buyers describe as a method. Before switching on an automated campaign type, decide what you will need to explain to leadership, and set up the reports and exclusions the platform offers before launch rather than after a bad month. Review recommended settings and automated creative options before accepting them, and test material changes against a comparison. Treat platform-reported conversions as attributed results, not proof of incremental sales: before reconciling them with your analytics, compare attribution windows, click and view-through rules, conversion definitions and any modelled conversions, and for material budget use incrementality or holdout tests to estimate what the advertising added. Judge a platform by your own order value and conversion rate against its auction, not by its rating. Ask who your support contact will be before committing budget. Audit which conversion events and values reach each platform, whether they are deduplicated and what modelling applies where observation is limited; consider server-side integrations where they fit, within your consent obligations; and give new campaigns enough time and conversion volume to judge, without assuming the learning period guarantees improvement. And before buying programmatic, name who will operate the DSP and how performance will be judged while it ramps up, ask for an itemised breakdown of media, DSP fees, agency compensation, reseller charges and other services, and compare the inventory spend with the total invoiced.
For how companies organise paid media and when a tool on top earns its place, see what paid media management software is; for the creative layer, what creative analytics software is; for the causal question, what incrementality testing is and what attribution-tool buyers wish they’d known.
What these interviews do not show
The ratings measure how satisfied each product’s own users are, not which platform performs best, and a company can rate a platform its agency runs. The incrementality findings are a few companies’ own tests and do not establish how much any platform over-reports in general. The support accounts are individual experiences and do not show which spend level earns a dedicated team. The moves described, such as one company’s switch of programmatic platform alongside a change of agency, are single cases, not a market trend. Product capabilities come from the vendors’ help pages and change often.
Common questions
Which ad platforms do buyers rate highest?
In Alium’s verified interviews Google Ads rates 7.8 out of 10, Display & Video 360 7.8, The Trade Desk 7.5, Meta Ads 7.4 and LinkedIn Ads 7.2. TikTok Ads rates around 7.5 and Microsoft Advertising around 6.5 on smaller samples. Among the tools that sit on top of the platforms, Motion rates 8.1 and Smartly around 7.5. A rating reflects how satisfied each product’s own users are, not a head-to-head test, and buyers rate the tool separately from what its auction costs them.
Are Performance Max and Advantage+ worth using?
Buyers in these interviews split. Some credit the automation with saving time and outperforming their own targeting: one luxury brand finds Advantage+ consistently beats the audiences it builds from its CRM data. Others say it has become a black box, which makes results hard to explain to leadership. Google’s help centre now lists a channel performance report, search terms and placement reports, negative keywords and brand exclusions for Performance Max; buyers describe the visibility that remains as partial, not an absence of those controls.
Can you trust the ROAS that Google and Meta report?
It is the platform’s attributed result, not proof of incremental sales: each platform credits conversions under its own rules, windows and signals. An apparel company whose incrementality tests found Google’s and The Trade Desk’s attributed revenue largely inflated is cutting back display and reconsidering parts of its search spend. Another buyer rates Meta a three because view-through conversions in Meta’s reporting do not reconcile with Google Analytics. View-through credit is where these buyers say the platform’s numbers and their own part ways.
Do you need a DSP like The Trade Desk or Display & Video 360?
Buyers who run one describe a platform that needs a dedicated owner, time to optimise and a clear view of fees. One rates The Trade Desk a six because it is complicated and needs someone owning it; another says Display & Video 360 takes a lot of training. Some reach Display & Video 360 only through a reseller that takes a percentage of spend, and one company moved its programmatic buying when its agency’s margins on the platform grew. Many buyers in these interviews run programmatic through an agency.
How good is support from Google Ads and Meta?
It depends on the account. Buyers with an assigned team describe it as a reason for a high rating, and one reaches Google through its agency’s premier status. Others describe slow responses and a hard-to-reach help process, including a large online retailer that finds support poor on both Google and Meta when something goes wrong. Two companies newer to TikTok praise the direct access they get to TikTok reps.
Get a checkup on your tech stack
A checkup for your marketing stack, built from hundreds of verified buyer interviews: what matters most to you, a shortlist for your stack, and your next steps.
Start my checkupFree · no password · your answers are anonymized before they inform research like this.