Categorized editor-related complaints — selection errors, unexpected state changes, copy-paste failures, and field positioning bugs.

Ranked and analyzed user-submitted requests with detailed descriptions to identify the highest-demand improvements.

Quantitative analysis of block type ratios, median fillable fields per document, and usage patterns across the ICP critical path.

Editor usability health score built from NPS data and customer journey mapping across key editor workflows.

Inherited recordings from a dedicated Usability Track covering block creation, field placement, selection, and multi-select workflows.

Consolidated dozens of prior research studies on individual editor components into a unified picture of systemic issues.

Cross-referenced usability findings, ticket clusters, analytics data, and session recordings to extract patterns the team couldn't compile from raw sources.

Customer Success reported users avoiding direct editing altogether — using uploaded PDFs instead of native documents. The editor was the gateway to PandaDoc's core value, and customers were finding ways not to use it.

For existing users the learning curve was so steep that some medium and larger organizations created dedicated internal positions just to operate PandaDoc's document editor on behalf of colleagues.

Without copy/paste or multi-select, users invented elaborate workarounds. To copy a block from page 2 to page 11: duplicate page 2 → move it next to page 11 in the page manager → drag-and-drop across. Users documented these flows and shared them internally.

Three squads actively modifying an aging codebase — one of the most complex in the product. Delivery time for new features was inflated by legacy constraints. Up to 70% of my time during active development was spent on support and ad-hoc adjustments mid-sprint.

Users with strong Word backgrounds consistently tried to select text across block boundaries — triggering block operations they didn't intend. The interaction model needed to make block vs. text editing visually unambiguous.

Users couldn't reliably identify whether an element was selected, focused, or in its default state. They described the editor as "not responding" when it was — they just couldn't read the feedback.

Users attempted multi-select in every session — it was a basic expectation, not a power feature. Its absence was a blocker, not an inconvenience.

Every element — blocks, fields, text blocks — now followed the same four-state model. Cursor type, border, and background shifted predictably at each stage. Users reported feeling "in control" for the first time in usability tests.

Replaced color-only field states with a combination of shape, label, and surface elevation. Designed a set of 8 accessible colors for multi-assignee workflows — replacing the original green/terracotta pair that was effectively invisible to colorblind users. All unassigned fields got solid transparent grey backgrounds so they remain readable even when overlapping content. Surface layering changes prevented fields from shifting when underlying text was edited.

Fields could now be assigned to multiple recipients in a single action. In usability testing every single participant asked for it and wanted to beta-test it before release. Batch operations — select, move, assign, copy, delete — replaced the previous 3–4 click-per-field workflow that had no group path at all.

The floating menu was redesigned in white — reducing visual prominence while emphasizing the selected block or field. Items were split into logical groups with enhanced hover and press states. Most-used actions (move, duplicate, delete) promoted. In Select state: all block-level options visible. In Focus state: menu shifts to text formatting for Text.block, with Enter/Escape navigation between levels.

All block and field properties — formatting, layout, field logic, assignment — consolidated into a unified right panel. Previously scattered across floating menus, a top toolbar, and hidden adjust icons. The right panel became the canonical reference: advanced options, conditional logic, and less-used settings all accessible in one place without hunting.

Snap guides became more responsive. Block structure is now preserved when copying and pasting within an established grid — a detail born from our own frustration as Notion users: tools that silently destroy grid layouts on paste force you to keep documents artificially simple. PandaDoc now keeps the full row/column/block hierarchy intact when extracting or inserting any block. About 8 states cover the full range of transfer principles between rows and columns.

Increased the selectable area for the entire block while preserving the ability to jump into text editing. A refined cursor model — arrow for block selection, I-beam for text — made the mode switch predictable without adding UI chrome.

The jump between block-selected and text-editing mode was previously instantaneous and jarring. Smoothed transitions made the experience feel more responsive and reduced disorientation — users stopped accidentally exiting text edit mode mid-sentence.

All new patterns were back-ported into the design system: token updates, component states, interaction specs. Every element affected by the editor redesign now had a single source of truth — ensuring consistency beyond the editor itself.