You are comparing two products, and you keep writing paragraphs that nobody reads. The reader wants to know where one wins and the other loses, screen by screen. Prose buries that. A structure that puts the same screen from both products next to each other, with a note on each, makes the difference obvious in a few seconds.
Here is a template you can copy for any two-product head-to-head. It fixes the comparison to three screens, pairs them row by row, and pins a comment to the exact spot that matters on each capture. Three screens is enough to show a real pattern and short enough that a busy reader finishes it. If you need the broader case for this approach, the product comparison use case walks through why pairing beats a feature table.
Pick the three screens before you capture anything
The screens are the spine of the whole thing, so choose them first. Pick three moments where the two products are doing the same job, not three screens you happen to have handy. A good default for most software:
- First meaningful screen: the signup, the empty state, or the first thing a new user lands on. This is where the product makes or breaks its first impression.
- The core action: whatever the product exists to do. Creating the thing, sending the thing, buying the thing.
- A moment of friction: pricing, an error state, a settings page, a required field. This is where products quietly differ and most comparisons skip it.
Adjust the three to fit what you are actually comparing. For two checkout flows, your three might be cart, shipping, and payment. For two dashboards, they might be the overview, a detail view, and a filter. The rule holds: the same job on both sides, or the pairing means nothing.
Capture both products on each screen, same frame
For each of your three screens, capture the same view from both products. Open the first product, click Capture screen, and grab the frame. Then open the second product, get to the matching screen, and capture it too. Try to frame them the same way so the reader is comparing like with like, not a full page against a cropped corner. If one capture is wider than the other, crop the still so both sit at a similar zoom.
You end up with six screenshots, two per row. In Cobalt Capture each screenshot becomes its own item, so the natural order is: Product A screen one, Product B screen one, then screen two for both, then screen three. The reader scrolls down and reads each pair as a unit.
If you already have screenshots, you do not need to re-capture anything. Paste them, drag them in, or use Add image. On a phone, where the browser cannot grab the screen, take a normal screenshot of each product and add it the same way.
Pin the one spot that carries the difference
A screenshot without a mark leaves the reader guessing what you meant. Drop a numbered pin on the exact element that makes your point: the field Product A asks for that Product B does not, the price Product B hides behind a click, the button Product A buries. One pin per capture is usually enough. Two if the screen genuinely has two things worth calling out. If you want a deeper look at placing pins so a reader reads them in the right order, the guide on pinning the exact spot a reviewer means covers it.
Then add the comment. Keep each one to a single claim the reader can check against the image: "Product A requires a credit card here; Product B does not," not "Product A's onboarding feels heavier." You can type the comment or dictate it with the browser's built-in speech recognition in Chrome or Edge. The pins and any arrows or boxes you draw are baked into the exported images, so whoever opens your comparison sees exactly what you marked.
The template, laid out
Copy this structure into your review. Each row is two items, screen matched, each with its own pin and note.
| Row | Product A capture | Product B capture |
|---|---|---|
| First meaningful screen | Pin + one claim | Pin + one claim |
| Core action | Pin + one claim | Pin + one claim |
| Friction point | Pin + one claim | Pin + one claim |
Add a short free-floating note at the top with your verdict in one line, so the reader knows where you landed before they scroll. A free-floating item has no screenshot, so it works as a plain header. Save the longer reasoning for the per-item comments, where it sits next to the evidence.
Publish it as a link your reader opens cold
On Publish, the review gets a short link. Anyone with that link can open it and read the pairs, no login and no app. There is also a clean read-only page for a reader who just needs to scan it, and PDF or Word exports if someone wants a file to keep. People you share the link with can comment on any individual item, so a teammate can add "we should steal this pattern" right on the capture that prompted it, and you can mark each comment resolved as you work through them.
Free to try covers unlimited reviews and screenshots with no credit card; after 7 days a review goes read-only for you, and the Pro plan at $5 a month keeps it editable with no cutoff. The shared link keeps working for your reader either way. If this comparison is part of a broader look at a rival, the competitive teardown use case extends the same pairing idea across a whole product.
Build one three-row comparison this week on two products you already know well. Once you see how fast a reader gets the point, you will stop writing comparison paragraphs for good.