FILTER ROAS 1.5 · SPEND > $50CROSS-REFERENCE GOOGLE TRENDSDOWNTREND → DRAFTEVERGREEN / UPTREND → OPTIMIZETAG BY MONTH · REVIEW IN 2 WEEKSFILTER ROAS 1.5 · SPEND > $50CROSS-REFERENCE GOOGLE TRENDSDOWNTREND → DRAFTEVERGREEN / UPTREND → OPTIMIZETAG BY MONTH · REVIEW IN 2 WEEKSFILTER ROAS 1.5 · SPEND > $50CROSS-REFERENCE GOOGLE TRENDSDOWNTREND → DRAFTEVERGREEN / UPTREND → OPTIMIZETAG BY MONTH · REVIEW IN 2 WEEKS
Icarus
Icarus SOP · Pythago feed optimization

Two feed optimization strategies.
One tool: Pythago.

The Icarus SOP for using Pythago to clean up your product feed: pull underperformers, cross-reference Google Trends, then draft the losers and optimize the winners.

01The SOP · Pythago walkthrough

Watch the two feed strategies. Then run the playbook.

→ Walkthrough · Pythago feed-optimization workflow (Feb release)

02The two strategies

Two feed moves inside Pythago: draft or optimize.

Strategy 1 — Analyze ROAS 1.5 and above.

Pull every product with a ROAS of 1.5+ and spend > $50 over 30 days. Cross-reference each one with Google Trends to decide: keep it live, or remove the underperformers fighting a declining market.

Strategy 2 — Optimize the stable & uptrend products.

For products on a stable or upward trend, don't draft them — optimize. Adjust price first (usually down ~$5), then enhance the product images. That's where 90% of the lift comes from.

Tag everything. Review in 2 weeks.

Across accounts, 20–30% of optimized products start performing within two weeks. The tag in Pythago is how you measure the lift.

03The Pythago playbook

The 7 steps to run both feed strategies end to end.

01

Filter Pythago: ROAS 1.5 and spend > $50 (30 days)

Open Pythago and set the filter at the top. You'll see every product that spent more than $50 in the last 30 days at a ROAS of 1.5 — your underperformers, with 30/14/7-day ROAS side by side.

Work at product level first, not variant level. Variants come later.
02

Scan the list — these were yesterday's winners

Most products in this list used to be bestsellers or had recent momentum. They're still pulling clicks and spend, but the ROAS has collapsed. Don't act yet — you need trend context first.

Compare the 30/14/7-day ROAS columns to see if it's a slow bleed or a sudden drop.
03

Cross-reference every product with Google Trends

For each underperformer, open Google Trends and check the curve. This single check decides the next move: downtrend → draft. Stable or uptrend → optimize. Never skip it.

Example: a cashmere turtleneck at ROAS 1.16 with a clear downtrend curve — that's a draft, not an optimization.
04

Downtrend → set the product to draft

If the trend is clearly down, select the product inside Pythago and set it to draft. It goes offline on the store. Trying to rescue a downtrend bestseller means fighting saturated markets with higher CPCs and fiercer competition — rarely profitable.

Protect profit, not ego. A drafted loser frees budget for the next winner.
05

Evergreen or uptrend → optimize the listing

If the trend is stable or up, the product deserves a second life. Two levers carry 90% of the impact: price first, then images. Need a faster image workflow? Use Icarus Product Image Hacks. Glance at SEO title and description — if the keywords are already in, don't waste time there.

Rule of thumb: drop the price by ~$5 if you can. If you recently raised it and killed momentum, roll it back.
06

Repeat for the ROAS 1.5 – 2.5 batch

Re-run Pythago with the filter set between ROAS 1.5 and 2.5 over 30 days. These aren't bleeding — they're underperforming. Same logic: stable/uptrend products go into the optimization batch.

Example: hearing aids on a stable-to-up trend — clear optimize candidate.
07

Tag everything in Pythago and review in 2 weeks

Tag every optimized product with the current month (e.g. "optimize-february") directly inside Pythago. Two weeks later, filter by tag and see which ones moved. Across accounts, 20–30% of optimized products start performing.

Without the tag you can't measure the lift. The tag IS the experiment.
04Before you close this page

The 7-line checklist.

Run the SOP once. Watch the margin move.

Re-watch the walkthrough whenever you need a refresher. The playbook above is the short version — bookmark it.