The paid pullback worked on the metric that matters: spend cut 31%, cost per subscription improved $201→$175 and combined ROAS rose 0.67→0.78 against the >1.0 KR. ChatGPT referrals more than doubled to an all-time record (647 sessions, 5.1% signup rate) while organic stabilized after five months of decline. New this issue: a Q3 OKR scoreboard up front, a fifth natural-language LLM query (95% baseline), and two watch items — an attribution regression bleeding paid signups into Unassigned, and Claude discounting RoomSketcher's own #1 ranking as vendor-sourced.
Read reportA root-cause analysis of the organic decline. Two separate Google events, five weeks apart: a June ranking demotion that cost impressions but almost no clicks, and an unannounced update on July 24 that cost 301 clicks a day — a quarter of US organic traffic. Roughly two thirds of that is ordinary ranking loss and recoverable; the rest is click-through falling on pages that still rank where they always did. Includes a prioritised page-by-page fix list with engagement data, a section-level breakdown of where the loss sits, and the tracking fault that currently prevents any sign-up being credited to a landing page.
Read reportFirst read against the new Q3 OKRs, and all three objectives start behind. Paid spend rose 28% but bought 4% fewer subscriptions (cost per subscription $171→$227) — traced to one campaign budget-capped at its best-ever CPA, then throttled. Non-brand US clicks fell 17.2% while rankings improved, so the loss is CTR compression, not demotion. LLM Visibility slipped modestly (87.5% / 95.8%) with RoomSketcher still #1 on ChatGPT, Perplexity and Copilot. Also corrects a Microsoft Ads currency error carried since Issue #1.
Read reportThe discovery story turned a corner: the LLM Visibility Score on "best floorplan software" jumped 79→96% (Gemini and Perplexity both flipped RoomSketcher to #1), head terms recovered ("floor plan software" onto page 1), and all-source sign-ups and purchases rose even as Google paid was cut. The new concern is paid efficiency — spend scaled ~56% (a Bing doubling) while conversions fell ~15%, so blended CAC rose from ~$35 to ~$64.
Read reportNew this issue: a 17-week trend view with line charts behind every headline metric. The story they reveal — Organic Search is in a structural -29% slide since February, not a blip. Plus: the site-speed fix regressed (mobile LCP back to 12.9s), and paid was deliberately reallocated across engines (P-Max paused on both; Bing conversions +24% as Google's eased).
Read reportA recovery cycle: the site-speed fix landed (mobile Lighthouse 41→66, LCP 13.6s→8.4s), paid scaled efficiently (combined CPA -29% to ~$35, Google conversions +76%), and sessions and sign-ups rebounded. The watch item: US organic softened sharply — impressions -27% and every head term slipped.
Read reportCombined paid New Subscription -46%, Google Branded NORAM conversions -40% on flat spend, head terms slipped — but the Floor Plan Creator page shipped, the Bing scale-back response worked, and ChatGPT US referrals doubled.
Read reportTwo-week pulse on US search visibility, AI citations, P-Max performance, and the next round of high-leverage moves.
Read reportFirst US-scoped pulse check. Free-creator keywords slipped, mobile score dropped 23 points, cross-engine AI picture got more nuanced.
Read reportInaugural external audit: search visibility, AI discovery, competitive landscape, and untapped keyword opportunity.
Read report