Lock in pricing, shipping, and messaging before code freeze hits Nov 1

Test depth to find where volume peaks and margin holds.

• Test 2-3 depths against your planned BFCM discount.
• Watch profit per visitor alongside conversion.
• Lock in your depth before BFCM week.

Ask Claude + Intelligems MCP

“Using my Intelligems data, show me my average discount per order and AOV for the last 90 days. Based on that, suggest 2-3 discount depths worth testing sitewide for BFCM.”

A strong headline lifts revenue without touching your margin.

• Test the headline your whole campaign points at.
• Run it long enough to clear a seasonal BFCM blip, not just a day or two of noise.
• Check whether the winner lifts conversion, order value, or both. 

Ask Claude + Intelligems MCP

“Using my Intelligems data, pull my sitewide conversion funnel for the last 90 days. Which stage has the biggest drop-off? Based on that, suggest a headline test, sale-led versus brand-led messaging, I could try on my highest-traffic landing page.”

Find the threshold that holds AOV without killing conversion.

• Keep your current threshold as control and test 1-2 higher variants. 
• Watch units per order, not just conversion, to see if people are adding items to hit the bar.
• Weigh the AOV gain against BFCM shipping costs.

Ask Claude + Intelligems MCP

“Using my Intelligems MCP, search my past experiments for anything on shipping thresholds or checkout. What won, and is it worth revisiting? Tell me how to relaunch it in Intelligems.”

Boost conversion with personalized offers for each segment.

• Start with signals you already have, like repeat purchasers or your loyalty list.
• Test a member-exclusive offer against your standard sale for that group.

Ask Claude + Intelligems MCP

“Using my Intelligems MCP, how does conversion compare between new and returning visitors on my site? Tell me how I could personalize each experience in Intelligems.”

Find the discount floor that converts these shoppers.

• Use Likelihood to Buy audience targeting for your Low visitors to determine what offer helps them convert
• Run a depth test against just that segment: 2-3 discount levels, watch where conversion actually moves
• Apply the winning depth only to that segment. Everyone else stays at your standard offer

Ask Claude + Intelligems MCP

“Using my Intelligems MCP, which discounts should I test for my low likelihood to buy audience?”

Protect margin on the categories that sell without a discount.

• Pick one category and test depth there before you touch the catalog.
• Run three depths to see where the return starts to flatten.

Ask Claude + Intelligems MCP

“Using my Intelligems data, pull revenue by product for the last 90 days and list my top collections. Which ones are currently being discounted the most, and which are selling well at full price? Based on that, suggest a collection where a different discount depth is worth testing.”

Bigger baskets without marking down your whole catalog.

• Most brands land on a free unit at the top tier, or 20–30% off the bundle.
• Show the tiers on both the PDP and in cart so shoppers see the next threshold before they check out.
• Set the first discount just above your average order size to boost cart size.

Ask Claude + Intelligems MCP

“Using my Intelligems MCP, pull my top 10 products by revenue this year. Which ones make sense for a BFCM bundle? Then tell me how I'd set that up in Intelligems.”

Create channel-specific landing experiences so your ad promise carries through to the site.

• Build a dedicated landing page for your BFCM paid social traffic.
• Match the headline and offer to the exact deal the ad promised.
• Split test it against your current page, but read results by channel, so BFCM traffic spikes don't mask what's actually working.

Ask Claude + Intelligems MCP

“Using my Intelligems data, compare conversion rate by traffic source channel for the last 30 days. Which channel is underperforming relative to its traffic volume? Based on that, suggest what a dedicated landing page for that channel should say and show.”