You control the campaign yourself
FroggyAds is self-serve. You fund the account, upload the approved creative, set the budget and choose sources yourself, so a one-change test stays a one-change test.
A 728x90 leaderboard is the wide, shallow Display banner you use when one brand cue, one offer and one action must be understood in a single left-to-right glance. We built FroggyAds as a self-serve platform so you can upload an approved leaderboard, choose eligible sources, cap the budget and judge every source by accepted customer actions, not by size or click rate alone. This guide shows you how to design, inspect and test the file before you spend.
Use a 728x90 leaderboard when the placement accepts the full 728-by-90-pixel canvas and your offer stays readable at that height. It is a strong first Display test when you have one truthful offer, one destination and a way to count accepted customer actions. On FroggyAds you upload the approved banner, choose eligible sources, set the budget and compare sources by the cost of each accepted action. If fitting your message into the strip would force text below a readable size, choose another supported composition instead of squeezing the copy.
Three questions decide it quickly:
A leaderboard test only teaches you something when you can hold the creative, destination and acceptance rule steady while you change the source or the budget. These are the platform features that make that practical for a media buyer, affiliate or agency.
FroggyAds is self-serve. You fund the account, upload the approved creative, set the budget and choose sources yourself, so a one-change test stays a one-change test.
Our network connects through 750+ SSP integrations, which gives you many sources to compare. Breadth is not a promise that every source carries 728x90, so confirm acceptance for the sources you intend to buy.
Delivery and spend records tied to a source let you set cost beside the accepted actions in your own system. That is the comparison that decides whether a placement deserves more budget.
Our platform integrates with Adscore and applies internal quality controls to help filter low-quality traffic. You can also use whitelists and blacklists to keep or exclude sources. None of this replaces your own accepted-action check.
The minimum account deposit is USD 50, and Display starting bids begin at USD 0.10 CPM. Both are starting figures, not forecasts of what your leaderboard will pay or deliver.
A 728x90 leaderboard answers one question: can a wide, shallow message earn a relevant visit without sacrificing legibility? Before you design anything, write down the audience, the page context you expect, the destination and the one customer action you will accept as success. Save the rendered file and the landing-page version with the test so you can explain the result later.
Keep the headline, brand cue, benefit and action short. Change one element only when the evidence points to a specific uncertainty. In supported inventory, the banner should remain distinct from nearby content without imitating navigation, alerts or controls.
A fixed 728x90 file does not adapt itself to other placements. Confirm the required size and delivery behavior first, then prepare separate approved assets when your media plan needs other shapes.
| Your situation | Leaderboard fit | What we suggest |
|---|---|---|
| One offer, a short brand name and one action | Strong | Build the strip and run a controlled source test. |
| Several products or price tiers | Weak | Lead with one product and move the rest to the destination. |
| A condition that must be visible in the banner | Risky | Shorten the offer or choose a format with more room. Do not hide the condition in a later animation frame. |
| Audience mostly on narrow phone screens | Check delivery first | Prepare a separate mobile composition accepted by the campaign. |
| An eligible known audience | Good | Restate one specific next step and link to the matching page. |
The width lets a designer separate brand, offer and action without stacking them. It does not justify extra claims, long conditions or several equal buttons.
At native scale, the viewer should understand the advertiser, the offer and the next step without searching across the strip. Give each role a clear place in the horizontal reading path, and remove competing messages before you reduce essential text size.
The destination completes the message. Keep the same offer and terminology on the landing page, and verify that important conditions stay clear before and after the click.
| # | Component | Operating requirement |
|---|---|---|
| 1 | Placement and device context | Capture the actual placement, surrounding content, device range and any scaling or crop behavior. |
| 2 | Exact dimensions | Build and export the required 728 by 90 pixel canvas without stretching another size. |
| 3 | Message hierarchy | Give brand, core message and action a clear reading order across the width. |
| 4 | Brand and imagery | Use one purposeful visual and keep essential brand elements recognizable at native size. |
| 5 | Copy and action | Write one accurate proposition and one next step; move secondary explanation to the destination. |
| 6 | File and animation rules | Verify the selected platform or publisher's current format, weight and animation requirements. |
| 7 | Accessibility and quality review | Check contrast, readability, identity, claims, rights and the final compressed export. |
| 8 | Variant and performance record | Give each asset a stable ID and connect placement, audience, destination and outcome records. |
A technically valid file can still fail when its hierarchy, page context or destination is wrong.
Our Display specifications include 728x90 among common sizes. Check the available source and device settings, then connect your approved creative to one relevant destination. You can control the initial allocation without treating a standard size as a promise of available volume.
Create My Free AccountConfirm that the buying destination accepts 728x90 and inspect where the banner will appear. Note nearby colors, controls and content density.
Create the canvas at 728x90 from the start. Do not rely on the ad system to correct a stretched, cropped or oversized design.
Decide what this impression should communicate. If the brief contains several benefits, rank them and keep the strongest.
Let the eye cross the banner without doubling back, and leave enough quiet space to separate each role.
Shorten the copy for the 90-pixel height and use imagery that still makes sense after export. Keep required conditions legible.
Apply the current rules for the actual file type and destination. Google Ads, for example, lists uploaded image and HTML5 requirements separately.
Review the final asset for readability, contrast, truthful claims, advertiser identity, rights and any policy requirements that apply.
Place the exported banner in a realistic page frame at native size. Check that it looks like an ad, remains intact and does not compete with nearby controls.
Choose FroggyAds when you want to test a leaderboard with identifiable creative, destination and source decisions. Use the same approved offer and acceptance rule across the comparison, then vary one meaningful layout choice. Our Display campaign controls let you manage the initial spend while your own measurement setup establishes the accepted customer outcome.
Review the delivery records you actually have, accepted outcomes and any independently available viewability report. Compare like-for-like source conditions and observation periods before refreshing the creative. A noisy day alone is not evidence of fatigue; look for a sustained change or a reproducible readability or destination problem.
A 728x90 leaderboard contains 65,520 pixels and has a 364-to-45 aspect ratio, about 8.09:1. That ratio favors a left-to-right reading path. In a hypothetical test, compare a direct offer-led layout with a brand-led layout while keeping audience, bid and destination constant. Decide from accepted conversions and source quality after matching the observation window.
| Measure | Definition discipline | Review cadence |
|---|---|---|
| Qualified response per viewable impression | Use only actions linked to a compatible measured-exposure population and one declared attribution window; otherwise report actions and viewability separately. | After enough downstream actions have matured |
| Viewability | The share of measurable impressions that met the reporting provider's viewability definition. | By placement during delivery review |
| Qualified CTR | Clicks or visits that meet the campaign's stated quality rule, divided by the chosen eligible impression base. | During creative review, with source context |
| Post-click conversion | The share of eligible post-click visits that complete the primary landing-page action. | After the conversion window has matured |
| File weight | The final exported asset size compared with the destination's current limit for that file type. | Before upload and after every export |
| Creative fatigue | A sustained decline for the same creative in comparable placements, reviewed with audience mix and delivery changes. | Across comparable time periods, not one noisy day |
Keep denominators visible. Click rate and session conversion rate answer different questions and should not be merged into one success claim.
A fixed 728x90 banner is a Display creative, not a Push or Pop asset. Use FroggyAds to select an eligible display campaign, record the creative version and compare its sources. A move to another format requires a separate composition and a separate test.
Create My Free AccountUse immediate brand and offer recognition when the placement appears before much page context. Keep the action distinct from the site's own navigation.
Assume the reader is focused on nearby content. Use a concise, relevant proposition that can be understood without interrupting the article's meaning.
For an eligible known audience, use the width to restate one specific next step and send the click to the matching page rather than a generic home page.
Long horizontal copy can still become unreadable at native size. Shorten the message instead of shrinking essential type or conditions.
The width is not permission to add another benefit or action. Protect the main scan path and move supporting detail to the destination.
Test the exported banner against likely page backgrounds. Check every text and action state, not just the largest headline.
Keep essential elements away from uncertain edges and preview any scaling or crop behavior documented for the actual placement.
Repeat the banner's promise and terms on the first relevant screen. A generic or contradictory destination wastes a clear leaderboard message.
When results disappoint, work from the earliest step in the chain. Fixing a later step first only hides the real cause.
| What you observe | Most likely area | First check |
|---|---|---|
| Impressions but almost no clicks | Delivered size, crop or readability | Inspect a delivered instance or placement preview at native size. |
| Clicks but no accepted actions | Offer continuity or destination | Compare the banner promise with the first landing-page screen and test the full click path. |
| One source performs far below the others | Source or page context | Compare cost per accepted action by source before editing the creative. |
| Strong start, then a sustained decline | Creative fatigue or audience mix | Compare matching periods and delivery changes, not a single day. |
Most defects appear before launch, when the export is reviewed at native size and the full click path is tested.
Display creative production needs a budget for the brief, copy, design, licensed assets, exports and native-size review. Add destination work, measurement and media to the campaign plan, and cost other required sizes separately. That makes the one-time preparation effort visible when you compare the cost of another leaderboard purchase.
Four different numbers are often blurred together. Keep them apart in your plan:
| Figure | What it means | What it does not tell you |
|---|---|---|
| Minimum deposit | The USD 50 needed to fund a FroggyAds account. | How much a sound test needs. |
| Starting bid | Display starting bids from USD 0.10 CPM. | The CPM a given source will actually charge you. |
| Actual media cost | The spend shown in your campaign reporting. | Whether the traffic produced accepted customers. |
| Test budget | The amount you choose so the test can reach its review point. | A guaranteed result. |
Hypothetical arithmetic, not a forecast: USD 50 at a USD 3 CPM buys about 16,667 impressions. At a 0.40% click rate that is roughly 67 clicks, and if 5% of those clicks became accepted enquiries you would see about three. Three events cannot separate two creatives with confidence. Set the budget from the number of accepted actions you need before you will decide, and treat a first deposit as an entry point, not a verdict.
A display campaign cost comparison needs the same audience question, offer and accepted outcome, with a review point and loss limit set before launch. Separate recurring media expense from the production work each variant requires. Plan review capacity as well as media spend, because new placements, versions and leads all create work.
Create your FroggyAds account with an approved leaderboard and a measurable destination. Set the spending limit for the first source comparison, retain the actual delivered-size evidence. Then review accepted customer actions before expanding the purchase.
Create My Free AccountSource-level measurement connects the paid leaderboard purchase with the customer outcome you can actually verify. Use FroggyAds delivery reports to retain source and spend, then match the supported references with qualified visits and later actions in your receiving system. Review that comparable cost and outcome evidence before giving a source more budget; delivery alone does not establish acceptance.
Before planning a leaderboard test, confirm current 728x90 inventory and creative acceptance for the sources you intend to use. Do not infer availability from the dimensions alone. When a source performs well, keep it; when it delivers impressions without accepted actions, use our whitelist and blacklist controls to narrow the next round.
Preserve the same offer from banner to landing page and final action. When copy, imagery, audience or bid changes, retain the prior approved version for comparison.
The useful tool makes the 728x90 file easy to inspect, reproduce and match with its campaign record. Before launch, read the final export once at normal scale: brand, proposition and action should be clear without zoom or a design-tool overlay. Then open the destination on intended devices, complete the primary action and save the approved asset, URL, tracking details and test label together.
A 728x90 banner, called a leaderboard, is a horizontal display ad 728 pixels wide and 90 pixels high, an aspect ratio of about 8.09:1. Use it when one brand cue, one offer and one action can be read left to right in a single glance. It suits wide placements and is a poor fit for catalogues, dense screenshots or several equal buttons. Check which eligible placements will receive the creative before you build a campaign around it.
Yes. 728x90 is among the common Display sizes we list, and you run it through our self-serve Display campaign flow. Availability still depends on the source, device settings and campaign, so confirm current acceptance for the sources you plan to buy. Our 750+ SSP integrations give you a wide supply base to compare, but supply breadth is not a promise that every source will carry a leaderboard.
The limit depends on the destination and the file class, not on the 728x90 dimensions. Google's uploaded-display specifications, for example, list a 150 KB limit for image uploads and treat HTML5 packages separately, but that is a Google requirement, not a FroggyAds-wide rule. Check the requirements shown in your campaign for the exact file type, and inspect the compressed export rather than the editable original.
Keep four numbers apart: the USD 50 minimum account deposit, Display starting bids from USD 0.10 CPM, the CPM you actually pay and your chosen test budget. The deposit funds the account; it is not a test plan. As arithmetic only, USD 50 at a hypothetical USD 3 CPM buys about 16,667 impressions, roughly 67 clicks at a 0.40% click rate. Set the budget from the number of accepted actions you need before deciding, then add design, destination and follow-up costs.
Give each zone one job: a brand cue on the left, one proposition in the middle and one action on the right. As a fictional layout, 16 px edges, a 100 px mark, a 428 px message and a 136 px button with two 16 px gaps total 728 px. Judge the compressed export at native size, check contrast against likely page backgrounds and move secondary detail to the landing page. If a required condition cannot stay legible, shorten the offer or choose another format.
Not safely. Scaling 728x90 proportionally to 320 pixels wide leaves a height of about 39.6 pixels, and a feature drawn 16 pixels high becomes roughly 7 pixels. That is geometry, not an upload specification. Keep the leaderboard for slots that accept it and prepare a separately designed, separately approved composition for narrower placements.
No. Click-through rate shows initial response only. Judge each leaderboard by cost per accepted customer action and post-click conversion, and where a viewability report exists, state its measured population. In our hypothetical comparison, a placement with a 0.40% click rate cost USD 15 per accepted enquiry while one with 0.30% cost USD 10, so the higher click rate was the more expensive route to customers.
Combine our controls with your own acceptance rule. FroggyAds integrates with Adscore and applies internal quality controls to help filter low-quality traffic, and you can use whitelists and blacklists to keep or exclude sources. Keep the creative and destination fixed while you compare sources, then judge each one by accepted actions. No filter can guarantee conversions or replace a review of your downstream results.
Work from the earliest fault to the latest. Confirm the delivered size, cropping and file loading first, then readable text at native size, then the page context, then offer continuity and the destination's conversion path. Keep available viewability evidence separate from assumptions about exposure, and change one condition per step so the next result explains the change. A single noisy day is not evidence of creative fatigue.
Scale once the served leaderboard is readable and produces accepted actions at a cost you can accept. Use our source-level reports to separate the sources behind the result from those that only deliver impressions, then raise one budget or source setting and judge the extra traffic separately. A successful leaderboard test does not prove that another size, device or placement will perform the same way, and paid traffic cannot guarantee sales.
These sources document Google-specific requirements and accessibility guidance. Verify the rules for your platform, publisher and file type.
Complete these prompts against the final exported file, not only the editable design, and store them with the campaign.
Describe the live placement, surrounding page context, intended devices, any scaling or crop behavior, and the date you confirmed which source accepted the 728x90 size.
Record the exact canvas, file type, weight, animation status, asset ID, imagery rights and the intended left-to-right order of brand, proposition and action.
State the one offer and action, any condition that must stay in the banner, and the tested destination URL with its tracking details.
Name the one layout change in this variant, the primary accepted outcome, the review point, and the native-size and click-path review results.
Launch your paid leaderboard test with a clear Display buying decision. FroggyAds gives you self-serve targeting, budget and source controls so you can manage the media allocation. Keep the approved strip, destination and accepted customer action identifiable, then use the resulting evidence to choose the next change.
Create My Free AccountA leaderboard gives you a long reading path, not a large reading surface. Its aspect ratio is approximately 8.09:1. That suits an advertiser name, a concise proposition and a next step arranged across a strip. It is a poor starting point for a product catalogue, a dense screenshot or several equal actions. Before producing variants, decide what the reader should notice first and which information must remain visible during a brief exposure.
Calculated scaling example, not an upload specification: reducing a 728x90 file proportionally to 320 pixels wide leaves a height of about 39.56 pixels. A feature drawn 16 pixels high in the original becomes approximately 7.03 pixels high. The arithmetic does not say what size your platform will accept. It does show why shrinking the artwork is not equivalent to designing a readable mobile banner. Keep the original leaderboard for a suitable slot and prepare a separately approved composition when the smaller placement requires one.
One fictional English-language layout could use 16 pixels at each outside edge, 100 pixels for the advertiser mark, 428 pixels for the message and 136 pixels for the action, with two internal gaps of 16 pixels. The widths total 728 pixels. This is a worked brief, not a required safe area or an optimal proportion. A longer brand name, another language or a material condition may need a different allocation. Adjust the composition rather than treating the example as proof that your actual wording will fit.
Inspect the message at its exported size. A headline that looks comfortable in a zoomed design window can fail after compression or placement scaling. Check the actual wording of the action and any price qualifier. If required conditions need more space, shorten the proposition or choose another eligible format. Do not hide the condition in a later animation frame or assume that the landing page excuses a misleading banner. The initial asset should communicate an accurate offer on its own.
A browser window 768 pixels wide does not necessarily provide 768 pixels of advertising space. In a hypothetical page with 24 pixels of padding on each side, only 720 pixels remain, so a native-width 728-pixel creative is eight pixels too wide. A buyer should ask for a placement preview or an inspected delivered instance and include the surrounding layout, rather than deciding from a tablet or desktop label alone.
Record the actual creative dimensions and the visible container separately. If clipping occurs, determine whether it comes from the placement, the wrapper or an incorrectly supplied file. Device targeting can limit where you buy; it cannot repair a publisher container. Our Display overview lists 728x90 as a common size, while eligibility and delivery still depend on the campaign. Retain the source identifier and the observed page context when requesting help with a rendering fault.
Google's uploaded-display specifications identify 728x90 as a leaderboard and separate image uploads from HTML5 packages. Its listed 150 KB image limit is a Google requirement, not a FroggyAds-wide limit. Use the requirements shown for your actual campaign and file class. A JPG, an animated GIF and a packaged HTML5 creative need different checks even when their outer dimensions match.
For animated work, inspect the opening, transition and resting state. Brand, offer and action should not depend on a viewer watching every frame. For a static export, check the compressed file rather than the editable original. Review text contrast against the artwork behind it; W3C contrast guidance also addresses images of text. Keep the approved filename, export settings and destination together so a later replacement can be traced to a real creative change.
Hypothetical records, not FroggyAds campaign results: a header placement delivers 40,000 impressions, 160 clicks and eight accepted enquiries for USD 120. A second placement delivers 20,000 impressions, 60 clicks and six accepted enquiries for USD 60. Both have a calculated media CPM of USD 3. The header's click-through rate is 0.40%, above the second placement's 0.30%. However, its media cost per accepted enquiry is USD 15 rather than USD 10.
| Fictional record | Header placement | Second placement |
|---|---|---|
| Impressions | 40,000 | 20,000 |
| Clicks | 160 | 60 |
| Click-through rate | 0.40% | 0.30% |
| Accepted enquiries | 8 | 6 |
| Spend | USD 120 | USD 60 |
| Media CPM | USD 3 | USD 3 |
| Media cost per accepted enquiry | USD 15 | USD 10 |
Cost per accepted customer action divides the relevant spend by outcomes meeting the declared acceptance rule. In the fictional records above, the higher click rate does not establish better acquisition economics. Neither placement can be called profitable without lead value and subsequent costs. Confirm comparable acceptance, attribution and observation windows before acting, and keep the small totals' uncertainty visible instead of treating one attractive ratio as permission for unrestricted spend.
Suppose, within those fictional records, the header has 32,000 measurable impressions and 16,000 viewable impressions. Its viewable rate among measurable impressions is 50%. If the second placement has 16,000 measurable and 12,000 viewable impressions, the corresponding rate is 75%. The unmeasured impressions are not automatically non-viewable. Keep the reporting provider's definitions and coverage alongside the rates, as explained in Google Ad Manager's viewability guidance.
Do not divide every attributed conversion by a viewable-impression total from a different population and call it a viewable conversion rate. Such a ratio needs compatible exposure records and attribution rules. Otherwise report accepted cost and viewability as separate measures. FroggyAds source reporting supplies the buying context. The existence of a source ID does not prove that every impression has a verified on-screen exposure record in your measurement system.
A header and a mid-article slot can accept identical creative dimensions but expose the message in different contexts. Capture position where it is actually documented by a placement preview, publisher information or an available report. Do not invent a page-position field in the FroggyAds interface. When position is unknown, name the comparison by the source information you do possess and leave that limitation visible.
Choose FroggyAds for a leaderboard test when you need to connect that creative evidence with practical spending and source decisions. Keep the same approved artwork and destination for a source comparison. Keep sources as consistent as practicable for a creative comparison. If several settings change together, describe the result as a combined campaign change rather than proof that the banner design alone improved performance.
Our pricing page distinguishes a USD 50 minimum account deposit from Display starting bids of USD 0.10 CPM. Neither is the actual media CPM in the fictional comparison or a promise of leaderboard availability. Budget separately for design, export checks and follow-up. Authorise the next media allocation for a specific question, such as whether the approved strip attracts suitable enquiries on an eligible source.
At review, keep the creative ID, destination version, source, spend and accepted-event definition together. Decide whether to continue, repair a rendering problem, restrict a source or prepare another leaderboard concept. That record makes our self-serve controls useful: you can implement a reasoned change and retain the earlier setup for comparison. Avoid widening the campaign merely because more impressions are available.
Start your 728x90 leaderboard banner campaign with FroggyAds