Skip to main content
These are the questions users ask most when they start using Superscale. For deeper explanations, follow the linked pages.

​
Context and setup

Put each piece of context at the lowest level where it is always true. Workspace is for rules shared by every brand. Brand is for identity, voice, guidelines, audiences, competitors, and ad accounts. Product is for one offer, SKU, app, category, or landing page. See Levels & sources.
Start with brand guidelines, product facts, customer language, approved ads, product shots, logos, and any claims or restrictions. If the output has to look a certain way, include visual references. See Reference images & winning ads.
No. Superscale is context-first, so durable facts should live in context. The prompt should say what you want now: audience, angle, format, references, and what variable to test.
Context files teach the agent what is true. Skills teach it how to run a repeatable workflow. Use context for brand facts, customer language, claims, and product info. Use skills for recurring reports, QA checks, research methods, or creative formulas.
Usually the agent lacks specific references or durable rules. Add approved winners, product shots, logo rules, customer language, and examples of what not to do. Then ask for one controlled generation.

​
Credits and troubleshooting

Generating assets costs credits. Chatting, planning, organizing context, and preparing the brief are where you should refine before spending. See How credits work.
If an iteration creates a new asset, expect it to use credits. Ask for surgical edits and lock what should stay fixed.
Downloading an existing output does not create a new asset. Generation is the credit event.
Report the failed generation or contact support. Superscale can inspect, retry, or refund credits when the failure is real. Do not keep regenerating the same broken prompt blindly.
Plan the script first, attach exact references, preview voice/pronunciation where possible, and start with a smaller test before rendering a full video batch.

​
Integrations and publishing

Not automatically. An integration can read data, import creative history, publish assets, or manage campaigns depending on permissions and platform support. See Integrations overview.
Yes. Read-only account data lets Superscale find winners, waste, fatigue, and creative patterns without changing anything in the account.
Publishing and budget actions should be approval-gated. Keep a human in the loop for spend, compliance, and brand risk.
Check account permissions, selected business manager, page access, and integration scope. If it still does not appear, send the account ID to support.
Yes. Manual export is useful for strict review workflows, unsupported platforms, or teams that prefer to upload through Ads Manager.

​
Research and creative strategy

Copy the structure, not the brand. Learn from hooks, pacing, proof, layout, and offer, then translate the pattern into your product and audience.
Add your own list, remove bad matches, and describe what makes a competitor relevant: same buyer, same price point, same market, same production quality, or same objection.
Public ad-library signals are directional. Runtime, variants, recurrence, and format repetition suggest what is worth studying, but they are not private ROAS.
Generate enough to test a hypothesis, not as many as possible. Change one variable at a time: hook, audience, value proposition, format, offer, or visual style.
Start with competitor research, reviews, and a small testing matrix. Once ads run, connect account data so future batches learn from your own results.

​
Video, voice, and localization

Not always. Many strong workflows are hybrid: AI for the hook or creator, real B-roll or screen recordings for product trust, and Superscale for captions, voice, and editing.
Think scene-by-scene. Regenerate or replace the weak scene where supported instead of rebuilding the entire video.
Add pronunciation guidance in brand context and in the video prompt. Use simple phonetics for names, acronyms, and industry terms.
Localize copy, visuals, audio, and market context. Do not only translate the script. Specify the market, accent, currency, and local examples.
Use real footage, less polished scripts, tighter scenes, a better voice, and more natural references. Hybrid workflows usually feel more authentic.

​
Products, assets, and formats

Yes, if that is the level the ad should reason about. Create a product-like entry for a collection or bundle and attach representative images and offer details.
Use a website, landing page, manual product setup, CSV, or attached product context. Shopify is the easiest path, not the only one.
Attach the exact product asset, keep the reference set narrow, and say what must not change: packaging, logo, UI, color, label, screenshot, or layout.
Superscale supports common paid-social formats. The export view shows which ratios are available for the creative type you are working with.
Use batch or project export controls where available. If a creative type still exports one by one, organize outputs in a project and download the final selections first.

​
Teams, agencies, security, and advanced workflows

Use one workspace with many brands when the agency owns operations and billing. Use separate workspaces when clients need isolated billing, members, ownership, or access boundaries. See Agencies.
Put durable legal, brand-safety, and claims rules into context. Use workspace level for agency-wide rules, brand level for one client, and product level for offer-specific restrictions.
Follow your contract, DPA, and internal policy. For unreleased or sensitive material, confirm the processing boundary before uploading.
Superscale for Agents is the direction for API and MCP workflows. Until a workflow is available in your account, use the UI and file handoff patterns.
Schedule narrow recurring jobs: weekly competitor scans, daily account health checks, Monday creative concepts, or report summaries. Define the source, cadence, and exact output.
Last modified on June 3, 2026