Ollav vs. screenshots in Slack
Most prototype feedback happens like this: someone posts a screenshot in a channel, people reply in a thread, and somebody tries to turn the thread into a to-do list. It works, and it costs nothing extra. This page compares it honestly with sharing the live prototype on Ollav.
Side by side
| Screenshots in chat | Live link on Ollav | |
|---|---|---|
| What reviewers see | A still image of one state | The working prototype: hover, click, scroll, type |
| What "this" refers to | Guesswork, or a drawn-on arrow | The element the comment is pinned to |
| Interactions and states | One screenshot per state, if someone takes them | Reviewers reach the state themselves; the clicks are recorded |
| Agreement | A reply per person | A reaction: 👍 5 |
| Knowing what's done | Re-read the thread | Open vs. resolved threads |
| The next iteration | New screenshots, new thread, context lost | New version at the same link, comments kept |
| AI agents | Can't read the thread without extra plumbing | Read open threads with get_review and resolve them |
| Setup | None | Upload a file or let an agent share it |
| Cost | Included in your chat tool | See pricing |
When screenshots are the better choice
- A quick gut check. "Which of these two headers?" is faster as two images in a channel.
- It isn't HTML. Figma frames, slides and PDFs aren't what Ollav hosts.
- Nobody needs to interact. If the question is about a static visual, a screenshot carries all of it.
When a live link pays off
- The prototype has flows or states, like onboarding, checkout or a wizard.
- Several people review it and you want agreement to count once.
- You'll go through more than one version.
- An AI agent built it and should act on the feedback itself.
Using both
Post the Ollav link in your channel. People still discuss it in Slack, and the specific, element-level feedback lands on the prototype, where the next version and the agent can see it.