×
GlobalIndiaThailandUAEUSA
Back
Back
Head of Marketing - Earned Media
Marketing | DV360
Improving CPM in DV360 is not about slashing bids and...
By Narender Singh
Feb 16, 2026 | 5 Minutes | 
| 
Look, anyone who been managing display campaigns in DV360 for more than a few months knows the drill. CPMs creep up, someone in finance starts asking pointed questions about efficiency and suddenly there pressure to cut costs. The knee-jerk reaction? Slash bids across the board and hope for the best.
That exactly the wrong move.
The real challenge isn't just lowering CPM. Any rookie can do that by tanking their bids. The trick is bringing down costs while your conversion rates, viewability and actual business metrics stay exactly where they need to be. It doable, but it requires thinking strategically rather than just pulling levers at random.
Here what usually happens. Lower your bids to improve CPM and your ads suddenly only show up on garbage inventory. You know the sites—the ones cluttered with pop-ups and questionable content that nobody actually reads. Sure, your CPM looks great in the report. Your conversion rate? Not so much.
The relationship between CPM and performance isn't actually antagonistic, though. What matters is finding inventory and targeting combinations where the impressions cost less but still reach people who might actually care about what you're selling. That means getting selective. Thoughtful. Not just competing in every auction because the inventory happens to be available.
Pull up your DV360 reports and segment by audience. Really dig into the data. What you'll probably find is that some of your most expensive third-party segments aren't actually pulling their weight. Meanwhile, that first-party list you built from website visitors? Might be converting at double the rate for half the cost.
Not all impressions are created equal and definitely not all audiences are worth the same price. Some segments cost a fortune because everyone bidding on them, not because they actually perform better. Once you identify which audiences genuinely drive results, you can double down there and pull back everywhere else.
It sounds obvious when you say it out loud, but you'd be surprised how many campaigns just target broad audiences and hope the algorithm figures it out. The algorithm good, but it not magic.
Site-level performance data is your best friend for CPM optimization. Go through your placement reports and you'll see patterns emerge fast. Some publisher sites consistently deliver strong performance at reasonable costs. Others? Expensive impressions that barely convert.
Build inclusion lists for the sites that work. Exclude the ones that don't. DV360 inventory packages make this easier than it used to be, letting you bundle up the good stuff and ignore the rest.
Here a truth that the premium publisher sales reps won't tell you: brand-name sites don't always perform better. Sometimes a mid-tier niche publication with an engaged audience will outperform a household-name publisher and it'll cost you 40% less per impression. Don't pay for prestige if prestige isn't driving conversions.
Private marketplace deals can be worth exploring too, especially if you negotiate based on performance guarantees rather than just haggling over price. The open exchange is fine for scale, but PMPs often offer better quality at more predictable costs.
With third-party cookies dying a slow death and privacy regulations tightening everywhere, contextual targeting has gotten both more important and more sophisticated. The upside for CPM optimization? Contextual often costs less than behavioral targeting while maintaining solid performance.
Use DV360 content category targeting combined with smart keyword lists. The goal is placing ads in relevant environments without being so hyper-specific that you drive up competition. Financial services ads in financial content. Travel ads in travel content. Revolutionary? No. Effective and cost-efficient? Absolutely.
Test contextual against your behavioral segments. You might discover that context-based targeting performs nearly as well at a fraction of the cost. Not always, but often enough that it worth testing seriously.
Showing someone the same ad seven times in three days doesn't make them seven times more likely to convert. It usually just annoys them. And you're paying for every single one of those impressions.
Check your frequency reports. Find the point where additional impressions stop generating incremental conversions. That typically somewhere between 3 and 5 impressions per week, though it varies by vertical. Cap your frequency there. The budget you save can go toward reaching new people instead of hammering the same folks repeatedly.
This is one of those optimizations that improves both CPM and performance simultaneously. Lower costs because you're buying fewer wasteful impressions. Better performance because you're not burning people out on your creative.
DV360 automated bidding strategies can help with CPM efficiency, but you have to set them up right. Maximum CPM bidding or Fixed CPM? You're probably overpaying. Those strategies just focus on winning impressions, not winning the right impressions at the right price.
Switch to Target CPA or Target ROAS bidding instead. Let the platform machine learning find the most cost-effective path to your actual goals. The algorithm will often pay lower CPMs for impressions that are more likely to convert, rather than competing aggressively across all inventory.
Don't strangle the bidding strategy with overly tight constraints either. If you set super narrow bid ranges right out of the gate, you're not giving the system room to find efficiency. Start broader, then tighten based on what the data tells you.
Better creative doesn't just improve conversion rates. It actually lowers your effective CPM too, because ads with higher engagement rates get preferential treatment in auctions. Think of it as programmatic version of a quality score.
Run creative tests constantly. Different formats, different messaging, different calls to action. Use DV360 creative rotation to automatically optimize toward winners. When one creative significantly outperforms the others, you'll start winning auctions at lower prices because the platform recognizes your ad generates better outcomes.
Also and this should go without saying but somehow doesn't: make sure your creative loads quickly and meets all the technical specs. Ads with file size issues or format problems perform worse and cost more. It basic stuff, but it gets overlooked constantly.
Programmatic inventory costs fluctuate throughout the day and week based on competition. Monday morning? Expensive, because everyone launching campaigns. Wednesday at 2pm? Usually cheaper. Sunday evening? Depends on your vertical.
Analyze performance by daypart and day of week. You'll often find that off-peak hours deliver similar conversion rates at significantly lower CPMs. Even if performance dips slightly, the cost savings can make it worthwhile from an efficiency standpoint.
Some brands get aggressive with dayparting and only run during their most efficient windows. If that Tuesday through Thursday afternoons for you, why compete during expensive Monday mornings? Test it. The savings might surprise you.
Different markets cost different amounts and perform differently. Breaking down campaigns by geography usually reveals some uncomfortable truths. That expensive metro market you've been targeting aggressively? Might have high CPMs without proportionally better performance. Meanwhile, some second-tier cities could be converting at half the cost.
Create separate campaigns or insertion orders for different geographic tiers. Bid more in markets that work. Pull back in markets that don't. Seems simple, but plenty of campaigns just target broad regions and treat every impression the same.
Brand safety matters. Obviously. But overly restrictive settings can limit your inventory pool and drive up costs without actually protecting your brand any better.
Review your current settings honestly. Are you excluding entire content categories that are actually perfectly safe for your brand? That limiting your inventory and making you pay higher prices for what left.
Same thing with viewability requirements. If you're optimizing for conversions and you've set a 100% viewability threshold, you might be unnecessarily restricting inventory. Test different viewability levels to find what actually impacts your end goals. Sometimes 70% viewability delivers the same conversion rate as 100% but at much lower CPMs.
The programmatic landscape shifts constantly. New inventory sources emerge. Audience behavior changes. Competition evolves. What worked three months ago might not work today.
Dedicate budget to structured testing. New inventory sources. Different audience combinations. Alternative bidding strategies. Use DV360 experiment features to run proper tests that give you statistically valid answers.
Document what you learn and scale what works. The campaigns that consistently deliver the best efficiency are the ones that evolve based on data, not gut feel or "best practices" someone read in a blog post two years ago.
There no magic bullet for improving CPM in DV360 while maintaining performance. Anyone who tells you otherwise is selling something. What works is a comprehensive approach that addresses multiple aspects of your campaigns at once.
Start with the areas where you're seeing the biggest inefficiencies. Overspending on audiences? Fix that first. Inventory quality all over the place? Focus there. The key is making deliberate changes and measuring their impact carefully, not just tweaking settings at random and hoping for the best.
CPM optimization is ongoing work, not a one-time project. Markets change. Competitors shift strategies. Audience behavior evolves. The campaigns that maintain strong performance long-term are the ones that continuously adapt based on what the data is actually saying.
With the right approach and consistent attention, you can absolutely improve your CPM in DV360 while keeping your performance metrics solid. It takes effort, but the payoff in campaign efficiency makes it worth doing right.