Conversion Optimization
Find where users drop off, which devices underperform, and which content groups deserve more investment.
MCPs required: SealMetrics MCP Best for: CMOs, Ecommerce Managers, Sales & Direct Directors
SEAL-006 — Weekly funnel drop-off alert
Using SealMetrics MCP for site {site_id}, compute the main conversion funnel week over week for the last 8 weeks.
Funnel steps (adjust if my actual funnel differs — ask me first if unclear):
1. Landing page entrance
2. View product / view room
3. Add to cart / start booking
4. Checkout / payment step
5. Purchase / booking confirmation
For each step return: sessions in step, % conversion to next step, week-over-week delta.
Alert me on any step where the step-to-step conversion dropped more than 15% vs the previous week. For each alerted step, give me the top 3 traffic sources that worsened in the same period.
Output: a 5-row × 8-week matrix + an "Alerts" section underneath.
SEAL-007 — Micro-to-macro conversion ratio
For site {site_id}, query SealMetrics MCP for the last 30 days.
For every micro-conversion type (add_to_cart, view_product, search, signup, view_room, etc.):
- Total micro events
- Sessions that fired the micro AND completed a macro conversion afterwards
- Ratio macro / micro
- Average revenue per macro that started with that micro
Rank micro-conversions by absolute volume of macros generated. Flag the ones with high volume but low ratio (high intent leak) and the ones with low volume but high ratio (under-exposed funnels).
Output: 1 table + a 3-bullet "what to amplify, what to fix" summary.
SEAL-008 — Top 10 landing pages and shared patterns
Using SealMetrics MCP for site {site_id}, return the top 10 landing pages by conversion rate in the last 90 days, with at least 200 entrances each.
For every page provide: URL, entrances, conversions, CR, revenue, top traffic source, top device, content group, average time on page.
Then analyze the list and extract 3 to 5 shared patterns (URL structure, source, content group, device skew, time on page). Phrase them as "rules of thumb" I can apply to other pages.
Final block: pick 3 pages from the rest of the catalog that violate these rules and propose A/B tests.
SEAL-009 — Mobile vs desktop conversion gap
For site {site_id}, query SealMetrics MCP for the last 30 days.
Per landing page (top 50 by entrances), return CR for desktop and CR for mobile, side by side, with the absolute gap and the relative gap.
Flag every page where mobile CR is more than 30% lower than desktop CR. For each flagged page, also pull bounce rate and time on page split by device.
Output: a sortable markdown table + a top-5 list of "mobile experience suspects" with a one-line hypothesis each.
SEAL-010 — Conversions by content group
Using SealMetrics MCP for site {site_id}, give me a full breakdown of conversions and revenue by content group for the last 90 days.
For every content group: entrances, conversions, conversion rate, revenue, AOV (revenue / conversions), share of total revenue.
Rank by revenue. Highlight content groups that capture more than 10% of entrances but less than 5% of revenue (overexposed losers) and content groups under 5% of entrances but more than 10% of revenue (underexposed winners).
End with a recommendation: which content groups should get more SEO/Ads/email investment next quarter.
SEAL-011 — Top SKUs per landing page (ecommerce)
For site {site_id}, query SealMetrics MCP using property breakdown on conversion_items.
For each of the top 30 landing pages by revenue last 30 days, return:
- Landing URL, conversions, revenue
- Top 5 SKUs purchased after landing on that page (sku, units, revenue)
Identify "landing-SKU" pairs where the top-converting SKU is NOT featured on the landing page (likely thanks to internal navigation). Suggest pinning that SKU above the fold.
Output: nested table grouped by landing page + a "missed merchandising opportunities" list at the end.
SEAL-012 — Cart-to-purchase ratio per category (ecommerce)
Using SealMetrics MCP for site {site_id}, for the last 30 days:
1. For every product category (property `category` on micro-conversion add_to_cart and on conversion_items): count add_to_cart events and count purchases.
2. Compute cart-to-purchase ratio per category.
3. Compute the abandonment delta vs site average.
Rank categories by absolute leak (add_to_cart events that did not turn into a purchase). Highlight the 5 worst.
For each of the top 5 leaks, suggest 2 likely causes (price, shipping, stock, payment friction) and what data point I should pull next to confirm.
SEAL-013 — High-intent search terms without dedicated landings
For site {site_id}, query SealMetrics MCP for the last 90 days.
List every utm_term (paid keyword) or organic search term with 3 or more conversions that did NOT land on a URL containing the term as a slug fragment.
For each term return: term, conversions, revenue, current top landing page URL, current CR.
Prioritize by potential lift (conversions × current CR vs site-best CR). End with a list of "landing pages to build" sorted by expected revenue impact.