A monthly Claude prompt that pulls GSC + SERP data for every vs-comparison page you've shipped, scores each page across 4 trajectory metrics, and tells you which to kill, refresh, double-down, or wait on. The measurement workflow that closes Track 02's diagnose-build-measure loop — every page that gets built also gets tracked.
A B2B SaaS team ships 9 vs-pages over a quarter using the gap finder + alternatives page system. Six weeks later, somebody asks the SEO question every CMO asks: "are they working?" The team checks GSC for total comparison-page impressions and sees a line that's slightly up. They declare victory. What they miss: 3 pages are at position #2-3 driving most of the traffic, 5 pages are stuck at position #15-25 contributing nothing, and 1 page got replaced by an aggressive competitor refresh and is regressing fast. The aggregate number hides the per-page reality.
The deeper problem is that vs-pages are a portfolio, not a campaign. Each page targets a specific query battery, climbs through Google's evaluation period at its own pace, settles at a position determined by its content quality + competitive context, and then either holds or drifts based on what competitors do over the next 12 months. A single quarter's batch of 9 pages will produce 3-5 winners, 2-3 stallers, and 1-2 regressors — and the right action is different for each.
This workflow runs the per-page assessment. Claude pulls GSC data for every vs-page in the portfolio, runs a SERP check via web search, scores each page across 4 trajectory metrics (average position, impression growth, CTR change, branded query lift), classifies each page into one of 5 trajectory states, and outputs a per-page action recommendation. Run monthly — the cadence catches stalling pages before they accumulate too many sunk costs and catches winning pages early enough to invest more behind them.
The gold variables — your brand, vs-page list with ship dates, and reporting period — are the parts you edit. Save the page list in a markdown file you keep updated as new pages ship.
google_search_console via the free Growthspree MCP for the GSC pulls, and Claude's built-in web_search for the SERP check. The portfolio list of vs-pages is the most important input — keep it updated as new pages ship via the Alternatives Page System workflow.Sample output for a hypothetical mid-market B2B SaaS brand with 9 vs-pages shipped over the last 5 months. The portfolio shows the typical distribution: 3 winners, 2 climbers, 2 stallers, 1 regressor, 1 too-new-to-evaluate.
Pulling GSC data for the 9 pages in your portfolio and running SERP checks for the 8 that are 6+ weeks old.
+ 6 more SERP checks
9 vs-pages assessed. Distribution: 3 WINNING · 2 CLIMBING · 2 STALLING · 1 REGRESSING · 1 WAIT (too new). Single most urgent finding: the Competitor B vs YourBrand page lost 7 position slots in the last 30 days after Competitor B shipped a refreshed comparison hub. Investigate before refreshing.
Build the portfolio markdown once, then run the workflow monthly with that file as the input. Pages added each month flow into the same review.
Compile a markdown file with every vs-comparison page you've shipped in the last 12 months. The ship date is the most important field — it determines whether each page is in the WAIT (under 6 weeks) or evaluation-ready (6+ weeks) bucket. Also include both directions where applicable (YourBrand vs Competitor AND Competitor vs YourBrand). Keep the file in the same Claude project so it auto-loads.
Edit the gold variables in the prompt — your brand, GSC property URL, and the portfolio list. Authorize GSC via the Growthspree MCP if not already connected (HubSpot is not required for this workflow). The GSC property must match the domain where your vs-pages live; subdomain or path-prefix mismatches will produce empty results.
For a portfolio of 8-12 pages, the workflow takes 5-8 minutes. Each non-WAIT page produces 1 GSC query + 1 web search. Pages classified as WAIT are skipped to save query budget. The output is the per-page status table + top 3 priorities — these are the two artifacts to focus on. The detailed metric breakdown is reference, not action.
Investigate REGRESSING pages first (always within the week), then refresh STALLING pages, then double-down on WINNING pages. WAIT and CLIMBING pages need no action — re-evaluate next month. Refresh actions should be passed to the Alternatives Page System workflow with the existing page as input. Two consecutive months of STALLING + a failed refresh attempt = KILL the page or repurpose the URL.
Refresh via Alternatives Page System →Same trajectory framework, different scope. Pick the one that matches your portfolio size and team.
Below 5 pages, single-page outcomes dominate the aggregate and trajectory states are noisier. Switch to a per-page deep-dive format that does fewer pages but goes deeper on each — including manual SERP screenshots, AI overview presence checks, and competitor change detection.
For brands with 12+ pages, the most actionable insight isn't per-page status but the *patterns* that separate winning pages from stalling pages. What headline structure, FAQ schema, and competitive positioning do your winning pages share that your stalling pages lack? Surfaces the production playbook.
For monthly internal use the per-page table is the right format. For quarterly executive reporting, roll up to a 1-page summary: total pages shipped this quarter, total impressions delivered by vs-pages, top 3 winning pages with traffic numbers, and one-line top priority for next quarter.
The portfolio is the work. Every vs-page you ship gets one of five trajectory states; the right action is different for each. Run the tracker monthly, refresh stallers, double-down on winners, kill the dead pages cleanly. Or have senior GrowthSpree operators run the monthly portfolio review across your vs-page portfolio, refresh stallers via the alternatives page system, and surface competitor regressions before they cost you SERP positions — the same operating motion run across 300+ B2B SaaS accounts.