Online Advertising Platforms: Where Budget Goes to Die

Online advertising platforms are the paid channels, including search, social, programmatic display, and video, that businesses use to reach target audiences and drive commercial outcomes. The major platforms include Google Ads, Meta Ads, LinkedIn, TikTok, Amazon Advertising, and programmatic networks such as The Trade Desk. Choosing the right combination depends on your audience, your funnel stage, and what you’re actually trying to accomplish commercially, not which platform has the best sales deck.

Most businesses don’t fail at online advertising because they chose the wrong bidding strategy. They fail because they allocated budget to platforms that were never going to reach the people who needed to hear from them, and then optimised relentlessly within that mistake.

Why Platform Selection Is a Strategic Decision, Not a Tactical One

Early in my career I spent a disproportionate amount of time optimising lower-funnel performance. Click-through rates, cost per acquisition, return on ad spend. The numbers looked good. Clients were pleased. And I genuinely believed we were driving growth. It took me longer than I’d like to admit to recognise that a significant portion of what we were “converting” was going to happen anyway. The intent was already there. We were capturing it efficiently, but we weren’t creating it.

That distinction matters enormously when you’re thinking about platform selection. Search advertising, particularly Google, is exceptional at capturing existing intent. Someone types a query, you appear, they click, they convert. The attribution is clean, the logic is intuitive, and the ROI looks compelling. But if nobody is searching for what you sell, or if the people who would become your best customers have never considered your category, search alone will not grow your business. It will harvest what’s already ripe.

The analogy I keep coming back to is a clothes shop. Someone who walks in and tries something on is substantially more likely to buy than someone who walks past. But the people trying things on didn’t appear from nowhere. Something made them aware of the shop, curious about the brand, willing to step inside. The upper funnel created the intent that the lower funnel captured. When you only invest in the lower funnel, you’re optimising the harvest while letting the crop thin out.

This is why platform selection is a strategic question. It forces you to ask: where in the buying process do we need to show up, and where are the people we haven’t yet reached spending their attention? Those two questions lead you to very different platforms than simply asking where you can get the cheapest conversion.

If you’re working through broader go-to-market decisions, the Go-To-Market & Growth Strategy hub covers the full commercial picture that platform strategy needs to sit inside.

What the Major Platforms Are Actually Built For

Every major advertising platform has a structural advantage that reflects how it was built and how users behave on it. Ignoring that structure is expensive.

Google Ads: Intent at Scale

Google’s core advantage is intent. When someone searches, they’re telling you exactly what they want, often with commercial signals baked in. “Best CRM for small business” is a very different signal from a passive scroll. Google Search is where you show up when people are actively looking. Google Display and YouTube extend reach into earlier funnel stages, but the core product is intent capture, and it remains the most efficient channel for that purpose in most categories.

The risk with Google is over-reliance. When I was running iProspect and managing significant search budgets across multiple clients, we saw a consistent pattern: businesses that had grown quickly on paid search hit a ceiling. The search volume for their core terms was finite. They’d captured most of it. Growing beyond that ceiling required reaching people who weren’t yet searching, which meant other platforms, which meant a different kind of investment with a different attribution profile.

Meta Ads: Scale and Social Context

Meta, across Facebook and Instagram, offers something Google doesn’t: the ability to reach people who aren’t actively looking but who fit the profile of someone who should be. The targeting depth, particularly interest and behavioural signals, combined with the creative canvas of visual and video formats, makes Meta well suited to building awareness and consideration at scale.

Meta’s challenge is signal quality post-iOS 14. The platform’s ability to optimise for downstream conversions degraded significantly when Apple restricted cross-app tracking. Advertisers who had built entire growth models on Meta’s attribution suddenly found their reported ROAS diverging sharply from what they could verify independently. The platform remains powerful, but it requires more sophisticated measurement thinking than it did five years ago.

LinkedIn: Professional Context at a Premium

LinkedIn is expensive on a cost-per-click basis. It is also, for B2B categories, often the only platform where you can reliably reach people by job title, seniority, company size, and industry simultaneously. That combination of professional context is structurally unique. The cost reflects the targeting precision, and for high-value B2B products, the economics often work despite the premium CPCs.

Where LinkedIn underperforms is in volume. If you need scale, the audience caps quickly. It works best as part of a multi-platform approach rather than a standalone channel, particularly for account-based marketing where you’re targeting a defined list of companies rather than a broad demographic.

TikTok and Emerging Video Platforms

TikTok and Short-Form Video

TikTok’s advertising proposition is built on attention. The platform’s algorithm is extraordinarily good at surfacing content to people likely to engage with it, which means well-crafted creative can reach highly relevant audiences at comparatively low CPMs. The catch is that “well-crafted” on TikTok means native-feeling content, not repurposed TV ads or static images. The creative bar is different, and brands that treat it as a cheap display channel miss the point entirely.

For consumer brands targeting under-35 audiences, TikTok has become a genuinely important awareness channel. For B2B or older demographics, the audience skew limits its utility, though that is shifting as the platform matures.

Amazon Advertising: Intent at Point of Purchase

For brands selling physical products, Amazon Advertising sits in a unique position. The intent on Amazon is not just commercial, it’s transactional. People searching on Amazon are not researching, they are buying. Sponsored Products and Sponsored Brands can intercept that intent at the moment of decision. For consumer goods brands, ignoring Amazon’s ad platform while selling on the marketplace is leaving a significant competitive disadvantage on the table.

Programmatic Display and Video

Programmatic, through platforms like The Trade Desk or Google’s DV360, allows advertisers to buy display and video inventory across thousands of publishers through a single interface. The advantage is reach and targeting precision across the open web. The disadvantage is brand safety risk, viewability variability, and the complexity of managing it well. Programmatic done poorly is a fast way to burn budget on low-quality impressions. Done well, it extends reach to audiences you can’t access through walled gardens at a competitive CPM.

How to Match Platform to Funnel Stage

The most common mistake I see in platform planning is treating the advertising budget as a single pool to be allocated to whichever channel has the best recent ROAS. That approach systematically underinvests in upper-funnel activity because awareness doesn’t produce clean short-term attribution, and it eventually starves the lower funnel of the new demand it needs to keep growing.

A more useful framework maps platform strengths to funnel stages:

Awareness: TikTok, Meta, YouTube, programmatic display, and LinkedIn for B2B. These platforms reach people who don’t yet know they need you. The KPIs here are reach, frequency, brand recall, and share of voice, not conversions.

Consideration: Meta retargeting, YouTube, LinkedIn, and Google Display. These platforms re-engage people who have shown some signal of interest, whether through site visits, video views, or content engagement. The creative should deepen the case for your brand, not just push for a click.

Conversion: Google Search, Google Shopping, Amazon Advertising, and Meta lower-funnel campaigns. These platforms capture active intent and should be optimised for efficiency. This is where ROAS and CPA metrics are most meaningful.

Retention and expansion: Email combined with paid social retargeting, and CRM-matched audiences across Meta and Google. These are often the highest-ROI activities because you’re marketing to people who already know you, but they’re frequently underinvested because they’re less visible than acquisition campaigns.

Understanding market penetration as a growth lever is useful context here. Reaching new audiences requires upper-funnel investment. Penetration-led growth doesn’t happen through conversion optimisation alone.

The Attribution Problem Nobody Wants to Solve Honestly

Every advertising platform reports its own performance in the most flattering light possible. That is not a conspiracy, it’s a structural incentive. Google will tell you Google drove the conversion. Meta will tell you Meta drove the conversion. If you’re running both, you will often find that the sum of attributed conversions across platforms exceeds your actual sales. This is the attribution overlap problem, and it is endemic.

I’ve sat in client meetings where the media plan looked like a success story on paper. Every channel was hitting its targets. ROAS was up across the board. And yet revenue growth was flat. The explanation, when we dug into it, was that the attribution models were cannibalising each other’s credit for the same conversions. We were measuring activity, not incrementality.

The honest solution to this is incrementality testing: running holdout groups to measure what would have happened without the advertising, and comparing that to what actually happened. It’s more complex than last-click attribution, it produces less flattering numbers, and it requires clients who are comfortable with ambiguity. But it’s the only methodology that gives you a defensible view of what your advertising is actually contributing.

Marketing Mix Modelling (MMM) is making a comeback for similar reasons. With signal loss from cookie deprecation and iOS restrictions, MMM provides a statistical view of channel contribution that doesn’t depend on user-level tracking. It’s not perfect, but it’s a more honest approximation than platform-reported ROAS at a time when cross-platform attribution is structurally broken.

The Forrester intelligent growth model touches on why measurement frameworks need to be tied to business outcomes rather than channel metrics. It’s a useful reference point for marketers trying to build a more defensible measurement approach.

Budget Allocation Across Platforms: What the Evidence Suggests

There is no universal right answer to how budget should be split across platforms. Anyone who tells you otherwise is selling something. What the evidence does suggest, from effectiveness research and from the pattern I’ve observed across hundreds of campaigns, is that businesses consistently underinvest in reach and brand-building relative to performance capture.

The long-term consequence of that imbalance is a shrinking pool of in-market buyers. You can optimise conversion rates indefinitely, but if fewer people are entering the consideration set for your brand, the funnel narrows over time. The performance numbers may look stable for 12 to 18 months while the underlying health of the brand erodes quietly.

A working principle, not a rule, is to think about budget in three buckets: investment in new audience reach, investment in re-engagement and consideration, and investment in conversion capture. The proportions depend on your category, your competitive position, and your growth stage. A new brand needs to weight heavily toward awareness. A mature brand in a competitive category may need to defend share of voice more aggressively. A brand with a large existing customer base may find retention and expansion the highest-return bucket.

The mistake is letting short-term attribution data make the allocation decision for you. Attribution models reward the last touchpoint, which is almost always a lower-funnel channel. If you let reported ROAS drive budget allocation, you will systematically defund the activity that creates future demand.

Scaling advertising investment effectively is covered well in BCG’s work on scaling agile operations. The principles around structured decision-making under uncertainty apply directly to how marketing teams should approach platform investment decisions.

Platform Costs, Benchmarks, and What They Actually Mean

Cost benchmarks for advertising platforms are widely published and almost universally misleading when applied out of context. Average CPCs, CPMs, and CPAs vary enormously by industry, audience, creative quality, bidding strategy, time of year, and competitive intensity in your specific auction. A benchmark from a report covering all industries tells you almost nothing about what you should expect in your category.

What benchmarks are useful for is identifying when your own performance is significantly out of range. If your CPCs are three times the category average with no obvious quality score or relevance explanation, that’s worth investigating. If your CPMs are half the benchmark, you should understand why before assuming it’s good news. It might mean your targeting is too broad and you’re reaching low-value audiences cheaply.

The more useful internal benchmark is your own historical performance over time. Are your CPAs trending up or down? Is your impression share holding or declining? Is your quality score stable? These internal trends tell you more about the health of your advertising than any industry average.

One thing I’ve learned from managing significant ad spend across multiple categories is that the platforms themselves are not neutral. Their optimisation algorithms are designed to spend your budget efficiently toward the objective you set, but “efficiently” is defined by the platform’s model of what drives the outcome you’ve specified. If your conversion tracking is imprecise, the algorithm optimises toward a proxy that may not reflect real business value. Garbage in, garbage out, but at scale and with confidence.

Creative as a Platform Variable

Creative performance varies significantly by platform, and not just in format terms. The context in which someone encounters your ad shapes how they interpret it. An ad seen in a professional context on LinkedIn carries different weight than the same message seen between cat videos on TikTok. The creative has to be native to the platform environment to work, which means your production approach needs to account for platform context, not just format specifications.

I’ve judged the Effie Awards, which are specifically focused on marketing effectiveness rather than creative craft. The entries that stand out aren’t the ones with the most polished production. They’re the ones where the creative idea is inseparable from the channel strategy. The medium and the message are integrated. That integration is rare in practice, particularly in larger organisations where creative and media are planned separately by different teams or agencies.

The practical implication is that your platform selection should inform your creative brief, not the other way around. Deciding where to run and then briefing creative to fit is a more effective sequence than creating an asset and then figuring out where to place it. It sounds obvious. Most organisations still do it the wrong way round.

Creator-led content is changing this dynamic on social platforms. Working with creators for go-to-market campaigns produces content that is inherently native to the platform because it’s made by people who live on it. The creative-platform fit problem largely solves itself when the creator understands the audience better than your brand team does.

When to Add a New Platform and When to Resist

There is always a new platform demanding attention. Pinterest, Snapchat, Reddit, connected TV, digital out-of-home, retail media networks. Each has a genuine use case for some advertisers and a compelling sales narrative for all of them. The discipline is knowing when adding a platform creates incremental reach or incremental complexity without incremental return.

A useful test is whether the new platform reaches a meaningfully different audience than your current mix. If you’re already reaching your target audience effectively through existing channels, adding another platform doesn’t grow your reach, it fragments your reporting and divides your team’s attention. The management overhead of running an additional platform is real. Campaign setup, creative adaptation, bid management, reporting, and optimisation all take time. That time has an opportunity cost.

The case for adding a platform is strongest when: the audience skew is genuinely different from your current mix, the format enables a creative approach that isn’t possible elsewhere, or you have evidence that the platform reaches people at a different stage of the buying process than your existing channels. Without at least one of those conditions, you’re adding complexity for its own sake.

Growth strategy, including channel expansion decisions, benefits from a structured framework. Understanding growth levers helps separate genuine expansion opportunities from tactical noise. The same principle applies to platform decisions: test with a defined budget and success criteria before committing to ongoing investment.

Building a Platform Strategy That Holds Up to Scrutiny

The discipline I try to apply when building a platform strategy is to start with the audience, not the channel. Where are the people I need to reach spending their attention? What are they doing when they’re on those platforms? What kind of message fits that context? Those questions lead you to the platform, rather than starting with a platform and working backwards to justify it.

From there, the strategy needs to account for the full funnel, not just the conversion end. That means being explicit about which platforms are doing awareness work, which are doing consideration work, and which are doing conversion work. It means accepting that the awareness investment won’t produce clean short-term attribution and making the case for it on different grounds: share of voice, brand search volume, new customer acquisition rate over time.

It also means building a measurement approach that sits above the platforms rather than relying on each platform’s self-reported metrics. That might be MMM, it might be incrementality testing, it might be a combination of both. What it shouldn’t be is a spreadsheet that adds up each platform’s reported conversions and calls that the total.

The growth loop model is a useful lens for thinking about how advertising platforms fit into a broader system of customer acquisition and retention. Platforms don’t operate in isolation. They feed and are fed by organic channels, word of mouth, product experience, and brand reputation. A platform strategy that ignores those connections will consistently overestimate the contribution of paid channels.

For a broader view of how platform investment connects to commercial growth decisions, the Go-To-Market & Growth Strategy hub covers the strategic context that individual channel decisions need to operate within.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions