MembraneMCP
Campaigns

Binary image survey (Yes / No)

Respondents see one image at a time and cast a simple Yes or No vote. Fast to complete, fast to analyze — best when you just need a read on whether something lands.

Best for: fast screeningMin 2 imagesScales easily to large sets

When to use it

Binary surveys are great when you want volume and speed. Each image gets a Yes/No vote in isolation — no comparison burden — so respondents move fast and you can test many more images than a round-robin allows.

Screening a large set

Got 20 candidate designs? Binary is the only format where that doesn't become a slog.

Testing appeal

“Would you click this?” “Is this for you?” “Does this feel on-brand?” All work well as yes/no questions.

Quick directional signal

Not trying to pick a winner — just trying to separate the hits from the misses before deeper testing.
The upload flow

Images go from your browser, not the AI

Same three-step handoff as every media campaign.

  1. Step 1
    AI sends you an upload link
    Describe the campaign and the AI generates a private upload page for you.
  2. Step 2
    You upload your images
    Open the link, drag files onto the page, and click “Done — Return to chat” when it's all up.
  3. Step 3
    You tell the AI you're done
    Back in chat, say “uploaded”. The AI verifies and creates the campaign.
Prep

What you'll need

  • At least 2 images — PNG, JPEG, or WebP, up to 10 MB each. Unlimited upper bound.
  • A campaign title, description, tagline, and category.
  • The yes/no questionrespondents will see above each image (e.g., “Would you tap on this ad?”).
  • A reward amount.
  • A landing-page link.
How it works

Describe the question. Upload. Go live.

Example conversation
  1. I want to run a Yes/No survey on 15 ad thumbnails for our app launch. Question: “Would you tap on this?” Category: mobile apps. Reward: 50. Call it “App launch — thumbnail screen”.
  2. Upload link ready — drop all 15 thumbnails on the page. Since you've got a big set, naming them something memorable (like “fire-hero” or “duo-chat”) will help a lot when you read the results later.
  3. All 15 uploaded and named. Done.
  4. Confirmed. Creating and starting the campaign. Each respondent sees all 15 images one at a time and votes Yes or No on each. Ask me anytime for a ranked breakdown.
Reading the results

You'll get a Yes-rate per image, ranked.

Ask your AI for an insight report and you'll see each image with its Yes percentage, highest to lowest — plus notes on which images respondents hesitated on longest (dwell time).

Try these prompts

Copy, paste, and tweak

Thumbnail screen

“Create a binary image survey called 'Thumbnail screen — Q2'. Question: 'Would you tap on this?'. Category: content. Reward: 40. Send me the upload link — I have about a dozen thumbnails to upload.”

On-brand or not

“Set up a yes/no survey on eight new product photos. Question: 'Does this feel on-brand?'. Category: brand. Reward: 50. Campaign name: 'Product photography — brand fit'. Turn it on once I'm done uploading.”

Screen ad concepts

“Run a binary survey on 20 static ad concepts. Question: 'Would you stop scrolling for this?'. Category: advertising. Reward: 60. Name: 'Social ad screen'. Send me the upload link.”

Get the ranking

“My 'Thumbnail screen — Q2' binary survey has been running for 48 hours. Show me the ranked Yes-rate for each thumbnail, and flag any where people spent a lot of time thinking before voting.”

Good to know

A few things worth knowing

2+
Minimum images
No cap
Upper bound
10 MB
Max per image
Instant
Goes live after upload

Scales cleanly to large sets

Unlike round-robin, response time grows linearly with image count — 20 images isn't much slower than 10.

No 'maybe'

Binary forces a choice. If you want nuance (“somewhat”, “it depends”), use a text survey with a rating question instead.

Live instantly after creation

No processing wait — campaign goes live the moment the AI turns it on.

Order is randomized per respondent

Each respondent sees images in a different order — first-image bias doesn't skew the Yes rate.
When binary is the wrong choice
If every image is decent and you want to know which is best, use round-robin instead — pairwise comparison surfaces subtle preferences that a yes/no vote can miss.
Next: Image sliderSee other campaign types