I run a small portfolio of browser tools and recently stopped treating deployment as success. The latest 28-day Search Console window showed 119 clicks and 19,368 impressions across eight sites, but the useful lesson came from three very different behaviors.
1. InvoicePad: visits without completion
InvoicePad had 20 active users in the last seven days, but the invoice_pdf_started event was still zero. Instead of adding more templates, I changed one product variable: a second Print / Save PDF action is now visible at the top of the live preview. The original editor button stays in place, and both placements report a source value so the next review can show which one works.
2. LivePhotoKit: specific file intent beats generic copy
LivePhotoKit's LIVP-to-JPG converter already had a real query signal: livp to jpg produced 57 impressions and 2 clicks. The current experiment changes only the search-result title to match that exact task. Download and share actions are tracked separately from page visits.
3. iWorkViewer: access is the job
iWorkViewer recorded 38 clicks and 3,110 impressions. The product is not trying to replace an office suite. Its job is simpler: help someone inspect a Pages, Numbers, or Keynote file on a device without Apple's desktop apps.
The operating rule
For the next 7?14 days I am not adding a ninth site, mass-producing pages, or changing several variables at once. The review will focus on completed tool actions, repeat visits, and whether the single change moved behavior?not whether another deployment returned HTTP 200.
Small tools become useful products when measurement changes what you build next.
Top comments (0)