Google Ads Video Campaign Bidding Strategies

Almost every conversation about google ads video campaign bidding strategies starts in the wrong place. People ask which option is "best", as if one setting were universally stronger. It isn't, and that framing leads to the most expensive mistake on the platform: choosing a strategy that optimises for something you don't actually want. A bidding approach is not a quality dial you turn up. It's a target you point the algorithm at, and if you point it at views when your business needs sign-ups, the machine will do exactly what you asked and exactly what won't help you.

So this isn't a feature tour. It's the decision I run, in order, every time I build a video campaign: what each strategy optimises toward, which objective it holds for, and where each one quietly breaks. I'll focus on the three that matter in practice — Maximize Views (the engine behind video view campaigns maximize views bidding), Target CPV and Target CPM — then the error that ties them together, which is optimizing for a view when the outcome you're paid on is an action.

The idea underneath everything below: the algorithm is extremely good at buying whatever you tell it to buy, and completely indifferent to whether that's what your business needs. Your whole job is to make "what you told it to buy" and "what you actually want" the same sentence. When they diverge, you get a beautiful dashboard and an empty funnel — and I've inherited plenty of accounts that were exactly that.

The three video campaign bidding strategies that matter

Strip away the naming churn and there are three google ads video campaign bidding strategies I reach for, each pointed at a different outcome. These are view-focused tools and live in a different family from the conversion-oriented Google Ads smart bidding strategies I use on Search and Performance Max, and that distinction is the whole point of this piece.

Maximize Views — buy the most views the budget allows

Maximize Views is the automation behind video view campaigns maximize views bidding: you hand Google a budget and it spends it chasing the largest number of eligible views it can find, adjusting the bid in real time. Because it automatically sets bids toward that single goal, it's the natural fit for a pure awareness push where the KPI genuinely is "how many relevant people watched this". Its strength is its narrowness — it does one thing relentlessly — and that same narrowness is its trap the moment your real goal is anything past the view itself. It's a top-of-funnel brand awareness instrument, nothing more.

Target CPV — control the average price of an engaged view

Target CPV lets you set the average cost per view CPV you're willing to pay and asks the system to hold to it while gathering volume. This is the CPV bid I use when I care about the cost discipline of engaged views on skippable formats and want a lever I can turn against delivery. It trades some raw volume for price control, which matters when a view is worth a specific amount rather than being valuable in the abstract. For the full picture of what a view actually costs across formats, I broke that down in my analysis of real YouTube advertising cost from live accounts.

Target CPM — buy reach at a fixed price per thousand

Target CPM flips the meter. You pay per thousand impressions at a price you set, whether or not anyone keeps watching, which makes it a reach instrument rather than an engagement one. It powers bumper and non-skippable buys, where the job is frequency and memory. Judging a Target CPM campaign by view logic is a category error, because it was never trying to earn the view — it was trying to be seen.

When Maximize Views bidding holds — and when it doesn't

Let me be precise about video view campaigns maximize views bidding, because it's the strategy people over-apply the most. It holds when three things are true at once: the objective is genuinely awareness, a view is a satisfactory outcome in itself, and there's no downstream action you're measured on. When I run a brand-launch flight and the brief is "get this creative in front of the right audience efficiently", it's the correct choice and excellent at the job. It gathers volume fast and keeps the price honest by construction.

Where it breaks is subtle, because it never looks like it's breaking. The campaign reports a huge view count and a low price, and every number is true. The problem is that Maximize Views has no idea what a conversion is, so it optimises toward the cheapest available views — and the cheapest views are, structurally, the least commercially interested viewers. Someone one click from buying is expensive to reach; someone idly autoplaying is cheap. Point the strategy at "most views" and it will systematically prefer the second person, because that's literally what you set.

I've inherited accounts where a Maximize Views campaign was the "best performer" on every video metric and had produced almost nothing the business could bank. The view rate was strong, the funnel underneath was empty. Nothing was misconfigured; the strategy was doing its job flawlessly. It was simply the wrong job. That's the pattern: Maximize Views failing looks identical to Maximize Views succeeding, right up until you look at what happened after the view — the same discipline I apply across all the TrueView ad formats I run.

Target CPV vs Target CPM: which holds for which objective

Once you're past "views for their own sake", the practical fork is between Target CPV and Target CPM, and the honest answer to "which is better" is that they answer different questions. Half the decision is made for you by the format, because the format largely determines which meter is even running.

Target CPV is the tool when you're buying engaged views on skippable in-stream, in-feed or Shorts placements and want to control the average price. Per Google's definition, a view on an in-stream ad counts when someone watches 30 seconds — or the whole ad if shorter — or interacts with it, with shorter thresholds on in-feed and Shorts. The exact criteria are in the official Google Ads cost-per-view bidding documentation. Because the viewer's choice to keep watching is itself a signal, this buys qualified attention: you pay for people who opted in, and the skip filters out the ones who didn't.

Target CPM is the tool when the objective is reach and frequency. You pay per thousand impressions, the exposure is guaranteed, the engagement is not. It's the right call at the very top of a funnel and the wrong call the moment you need to know whether anyone cared. The rule I use: if the viewer's decision to keep watching is information I want to buy, Target CPV; if I only need eyeballs and I'll measure by lift, Target CPM. I pick the format for the objective first, and that choice hands me the meter — the bidding decision should be the last one, not the first.

The error: optimizing for views when you want action

This is the mistake the whole article is built around: if you want action but bid on views, the algorithm optimises for the wrong thing, and it will succeed at the wrong thing. Every view-based strategy — Maximize Views, Target CPV, Target CPM — optimises for a view or an impression. None optimise for a sign-up, a lead or a sale. So when someone tells me "I ran YouTube and got tons of views but no sales", I know what happened before I open the account. They asked for views and got views.

The logic is unforgiving. The algorithm finds the cheapest path to the objective you set. If the objective is views, the cheapest path runs straight through the least interested audience, because those views are cheapest to buy. The more efficiently it runs, the deeper you dig the hole. This is the same trap as importing a cost per click instinct from Search: on Search a click is a fairly direct intent signal, so a low cost is often good; a low view price with weak retention buys you far less.

What to do instead when the goal is action

When the objective is action, the strategy has to be pointed at the action, not the view. On video that means a conversion-oriented setup, so the signal the algorithm chases is the thing you're paid on. For lower-funnel work I lean on TrueView for action, built to drive clicks and conversions rather than raw views. Those conversion approaches use machine learning to optimize toward outcomes — the same family as Maximize conversions and Target ROAS on Search, where the system chases return on ad spend rather than exposure — and they replace the older reflex of manual bidding with signal-driven automation. Switch the objective and the same budget starts buying more expensive but far more valuable attention.

None of this makes the view worthless when you're chasing action. It's a superb diagnostic: a weak view rate on a conversion campaign tells me the creative isn't earning attention, and I'll fix the hook before I touch anything. What it must not be is the thing the auction optimises for. Keep the view as a health check and put the action at the centre, and you've resolved the most expensive confusion in video campaigns. A strong video ad still needs the right target behind it.

How I actually choose, in order

The decision only looks complicated when people make it in the wrong sequence — reaching for a setting first and reverse-engineering an objective to justify it. I go the other way. First, I name the outcome I'm actually paid on, in one sentence, with no hedging. "I want this creative watched by the right audience" is a different sentence from "I want demo requests", and admitting which one it is settles most of the downstream choices.

Second, I pick the format that serves that outcome — skippable engagement for qualified views, bumper or non-skippable for guaranteed reach, action formats for conversions. Third, and only third, I choose the bid strategy, which by now is usually the single option matching both outcome and format. A view outcome doesn't need me to set the highest possible price; it needs a bid that clears the auction and no more. Video isn't the same as chasing absolute top position in Search or optimising toward a target impression share — a video views campaign is judged on whether the right people watched, not on where a text ad ranked in the google search results.

Then I let it learn, then I read the whole chain

Once it's live, I resist judging it on day one. Video strategies need signal to settle, so I give the campaign room to clear its learning phase before reading anything into the numbers. When I do evaluate, I never grade on the strategy's own headline metric alone. A Maximize Views campaign with a gorgeous cost per view still gets read against what happened after the view; a conversion campaign gets read on cost per action. I compare each campaign to itself over time, never to a benchmark from someone else's account.

That's the whole discipline, and it's less about clever settings than honesty at step one. Choose the objective truthfully, let the format follow, let the strategy follow that, and point the algorithm at the outcome you're genuinely paid on rather than the one that's cheapest to buy. Unlike the broader google ads bidding strategies that chase conversions directly, video is judged on watched attention — so real google ads success on video comes from that alignment, not from a magic setting — and the anxious hunt for the "best" strategy quietly disappears, because there was never a best one, only a right one for the job in front of you.