Google display network cost has no fixed rate card, because the Display Network sells ad space through a real time auction every time someone opens a page or an app with an ad slot. What you actually control is the daily budget, the bid strategy and the quality of the creative. The unit price you end up paying is an output of that auction, not a number you set in advance.
This guide covers when money leaves the account, how to size a starting budget that produces usable data, why the Thai CPM benchmark tables floating around the web should not be trusted, and how to stop spend from drifting using controls that already exist inside the account.
Does the Google Display Network charge CPC or CPM
Both are possible, and which one applies depends on the bid strategy you pick inside the Display campaign rather than on the network itself. A common assumption is that Display automatically means CPM, which is not the case for most accounts running on sales or lead goals.
- Click based strategies such as Maximize clicks or Manual CPC charge you when somebody clicks the banner. Impressions that get no click cost nothing.
- Conversion based strategies such as Maximize conversions or Target CPA still charge per click. The difference is that the system bids harder on people who look likely to convert and accepts a higher price per click for that group.
- Viewable CPM, or vCPM, is used for awareness campaigns. You pay for every thousand impressions that count as viewable.
- Target CPM appears in reach oriented campaign types, where the system works to hold your average cost per thousand impressions near the target you set.
Viewable has a precise technical meaning at Google. A display banner has to have at least half of its pixels on screen for at least one continuous second, and a video ad for two seconds. An impression that loads below the fold and never gets scrolled into view is not counted as viewable and is not billed under vCPM. That is why impressions and viewable impressions differ in your reports, and why the two numbers must never be mixed inside the same calculation.
The practical summary for a new account: if the goal is sales or leads, start on a strategy that bills per click, because you pay only when somebody was interested enough to tap. Keep vCPM for the moments when you are deliberately buying visibility, such as a product launch or a seasonal push measured on reach rather than on closed sales.
How much budget should you start Display Ads with
The honest answer is that there is no standard figure that works across businesses, but there is a way to work backwards from what you need to know when the test ends. A good starting budget is one large enough to produce a decision inside two or three weeks, not the smallest amount the platform will accept.
The calculation has four steps.
- Decide what you are measuring: people reached, clicks to the site, or conversions.
- Convert that outcome into the number of units you have to buy, meaning impressions or clicks.
- Multiply by an assumed unit price at first, then replace that assumption with your real number after two weeks of data.
- Divide by the number of days you plan to run, which gives the daily budget you type into the campaign.
The worked example below uses invented round numbers purely to demonstrate the method. They are not market rates for Thailand and should not be quoted as such. Suppose a CPM of 40 baht and a goal of reaching 100,000 people at an average frequency of 3 impressions each. That means buying 300,000 impressions. Divide 300,000 by 1,000 and multiply by 40 and you get 12,000 baht of media. Spread over 30 days, the daily budget is 400 baht.
Here is the same method on the click side. Suppose a CPC of 3 baht and a goal of 3,000 clicks in order to build a pool of visitors you can retarget later. The media budget is 9,000 baht, and over 21 days the daily budget lands around 430 baht. Again, the 3 baht figure is invented to show the formula, and it is not an average for any category.
One more thing to know about daily budgets: Google does not spend exactly the same amount every day. It can go over the daily budget on days with good traffic, while keeping the monthly total within your daily budget multiplied by the average number of days in a month. So when you see one day come in above the number you typed, do not immediately cut the budget. Check the running monthly total instead. Cutting budgets mid flight also forces automated bidding back into learning, which costs both time and money for no benefit.
The other budgeting mistake worth avoiding is splitting a small amount across many campaigns at once. An account spreading a few hundred baht a day across five campaigns ends up with five thin data sets that prove nothing. Concentrating the money in one campaign until a clear signal appears, then splitting, gets you to an answer faster.
What is the average GDN CPM in Thailand
Google does not publish official average CPC or CPM figures by industry for Thailand. The benchmark tables you see in articles are samples pulled from one third party provider's book of accounts, which reflects the categories they happen to serve, the budgets involved, the period covered and how they chose to define each industry. They cannot be used as a target for your account, and they should not sit underneath a quote you hand to a client or a board.
The workable alternative is to derive your own number in two stages.
Stage one, before you launch. Use the estimate Google Ads shows while you build the campaign. When you enter a daily budget in the Display campaign setup screen, the system projects a weekly range for reach and impressions. That range moves with the targeting you select and should be treated as a rough frame for what the money buys, not as a quotation. Keyword Planner forecasts prices for Search only and does not forecast Display CPM, so do not substitute a Search CPC estimate for a Display one.
Stage two, after two or three weeks live. Open the campaign report, use the columns button to add cost, impressions, clicks and average cost per thousand impressions. You can also compute it yourself: cost divided by impressions, multiplied by one thousand, gives your real CPM. Do the same for CPC by dividing cost by clicks, then break both down by device and by placement. The gap between mobile app inventory and desktop web inventory usually shows up immediately.
Once you have your own number, keep it as the month over month baseline for your account. That is the only benchmark that carries meaning, because it reflects your targeting, your creative and your landing pages. None of it holds up if measurement is broken underneath, so if you are unsure whether the events in GA4 line up with the conversions in the ads account, settle that before drawing conclusions about whether the price is high or low.
Why Display is cheaper than Search but converts less
Because the two networks buy completely different moments. Search buys intent that the user announced by typing a query. Display buys the attention of somebody reading the news, watching a clip or playing a game, and the ad interrupts that activity. Cost per click on Display is far lower because the supply of ad space online is enormous and competition per slot is lighter, and the downstream results are lower in the same proportion.
The problem this creates is one of reading the numbers. If you judge Display on last click alone it will always look poor, because someone who saw a banner today may search your brand name four days later, at which point the whole credit lands on the brand Search campaign. A fairer read looks at assisted conversions, at branded search volume moving in step with Display flighting, and at the return rate of previous visitors.
The other frequent trap is pouring Display budget into broad targeting from day one and then concluding that Display does not work. A safer order is to start with people who already know you, meaning remarketing and existing customer lists, where cost per conversion is usually the best in the account. Once measurement is stable, expand towards audiences who have never heard of the brand. Comparisons against Search campaigns should be made on the role each plays in the customer journey rather than on raw channel to channel CPA.
What banner sizes do you need to prepare
There are two routes. The first is responsive display ads, where you upload the parts and the system assembles the banner. The second is fixed size banners that you design yourself at every dimension. For a new account, the first route covers far more inventory for much less design work.
These are the assets responsive display ads ask for.
| Asset | Recommended size | Minimum size |
|---|---|---|
| Horizontal image, 1.91:1 (required) | 1200 x 628 px | 600 x 314 px |
| Square image, 1:1 (required) | 600 x 600 px | 300 x 300 px |
| Square logo, 1:1 | 1200 x 1200 px | 128 x 128 px |
| Horizontal logo, 4:1 | 1200 x 300 px | 512 x 128 px |
You can upload up to 15 images per aspect ratio and up to 5 logo files. On the text side the character ceilings shape how you write. Short headlines run to 30 characters with up to 5 versions, the long headline runs to 90 characters as a single asset, descriptions run to 90 characters with up to 5 versions, and the business name is capped at 25 characters. Up to 5 videos can be added, with around 30 seconds as the length Google suggests.
If you go the fixed banner route, the sizes that reach the most inventory and deserve to be built first are 300 x 250, 336 x 280, 728 x 90, 300 x 600, 160 x 600, 970 x 250, 320 x 100 and 320 x 50. Those last two are the mobile placements, which matter a great deal in Thailand. Uploaded image files must stay under 150KB, in GIF, JPG or PNG, and an animated GIF has to run 30 seconds or less.
The mistake that quietly wastes budget is shipping only one size. The campaign can then only enter auctions for slots matching that dimension, so impression volume stays oddly low while budget sits unspent. When time is short, run responsive first and add fixed banners later for the placements the reports show are working.
Controlling spend with frequency capping
Display budgets rarely run away at the total level. They run away by paying repeatedly to reach the same small group far more often than needed. Frequency capping in Google Ads sets a ceiling on how many times one person sees your ads inside a defined window. You set it in a Display campaign under additional settings, choose frequency management, then specify impressions per day, per week or per month and whether the cap applies at ad or campaign level.
A practical way to set it: read your current average frequency report first, then place the ceiling below what is already happening. A cap set above the frequency that actually occurs does nothing at all. On remarketing campaigns aimed at small pools, frequency climbs much faster than on broad campaigns because the same budget chases fewer people, and that is exactly where a cap saves the most money.
Alongside the cap, a few other controls belong in the same setup session.
- Shorten membership duration on remarketing lists, for example from 540 days to 30 days for products with a quick decision cycle.
- Exclude irrelevant placements, particularly mobile game apps, which tend to absorb large volumes of impressions through accidental taps.
- Exclude recent purchasers so you stop paying to sell to somebody who paid you yesterday.
- Keep budgets at campaign level rather than shared across campaigns while you still do not know which campaign earns its money.
Placement exclusions should be a weekly routine rather than a one off setup task, because new inventory joins the network continuously. Opening the placement report and cutting whatever spent money without producing results is one of the highest return per minute jobs in running a Display account.
How GDN and programmatic pricing differ
They differ in cost structure, not only in the headline cost per thousand impressions. Buying through Google Ads, what you pay is the winning auction price, with no separate platform fee layered on top. The inventory you reach is Google's network plus the publishers selling through Google's systems.
On the programmatic side, buying through a demand side platform, costs stack in layers. Typically that means the winning media cost, a platform usage fee, data fees for audience segments bought from providers, verification fees for brand safety and viewability measurement, and the management fee of whoever operates the platform. The result is that the amount reaching the publisher is lower than the amount leaving your account. What that gap buys is something the Display Network cannot offer: buying across multiple exchanges, negotiated direct deals with publishers, and deeper reporting control.
A quick decision rule. If monthly budget is modest and the goal is measurable sales or leads, staying inside Google Ads is usually better value because no fee layer sits on top. If the budget is large, you need specific inventory chosen by name, or brand safety requirements are strict, then paying fees in exchange for control starts to make sense. The important point is to avoid comparing the two CPM figures directly, since one includes fees and the other does not.
Agency fee models deserve the same scrutiny. Some are a percentage of media spend, some a flat monthly retainer, some tied to results, and some a mix. Each creates a different incentive, so ask early what the fee covers and whether creative production sits inside or outside it.
What this means in the Thai market
Thai traffic skews heavily to mobile, which means most impressions in a Thai account come from in app and mobile web inventory. Two consequences follow. Mobile banner sizes are mandatory, and app exclusion lists need real maintenance, because accidental taps in game apps inflate click counts without producing anything.
Thai language creative carries a visual constraint designers often forget. Thai script stacks tone marks and vowels above and below the baseline, so text that reads cleanly on a desktop monitor turns into a smudge once the image is scaled down into a 320 x 50 slot. The fix is to put as little text as possible into the image and let the words that must be read live in headlines and descriptions, which the system renders as real text at any size.
One more item to check before drawing conclusions about cost is the billing side. The budget you set in the account is media spend, while any tax added on top appears on the Google invoice according to your company's registration status. Budgeting without accounting for that is why the number reported upwards and the number the finance team sees fail to match every month.
Finally, Thai users spend most of their attention inside social and chat platforms, which means the sensible role for Display in most Thai accounts is following up and building familiarity rather than closing the first sale. Setting that expectation early leads to better budget decisions in month one.
FAQ on Google Display Network cost
Can Display work on a very small budget
Yes, provided you keep the audience narrow, for example only people who visited the site in the last 30 days. A small budget spread across broad targeting is too thin to produce any effect, while the same money aimed at previous visitors builds enough frequency for people to remember you and come back.
Why does some days' spend exceed the daily budget I set
Because Google is allowed to overspend the daily budget on days with better opportunity, offsetting it on slower days, while keeping the monthly total within the daily budget multiplied by the average number of days in a month. Check the monthly running total rather than reacting to a single day.
Does a cheaper CPM mean a better campaign
Not necessarily, because CPM often falls when the system buys lower quality inventory such as apps with frequent accidental taps. Read CPM next to viewability rate, next to the share of sessions that do not bounce instantly, and next to the final outcome before deciding cheaper is better.
Should I use fixed banners or responsive display ads
Start with responsive display ads because they cover far more inventory for less design work, then produce fixed banners only for the sizes and placements your reports show are performing.
How many weeks before I know whether the price I am paying is reasonable
Around two to three weeks of steady spend without major setting changes along the way, because automated bidding needs time to learn and changing budgets or goals mid flight resets that learning.
Wrapping up
The price of Display sits less in the unit rate and more in who the money reaches, at what frequency, and whether the measurement behind it is sound. If you want a team to structure the campaigns, set the frequency ceilings and get measurement into a state where the reported numbers can be trusted, Relevant Audience handles that work inside its Google Ads service. Get in touch with your goals and the budget you have to work with.
