IG volume is up — but it's cheaper, Reels-heavy traffic that converts far worse.

The brief — for the media team

Situation. Instagram's share of site traffic has climbed toward ~50%, but its conversion rate has fallen roughly a third — from 2.2% to 1.5% (orders per visit) year-over-year, and it's still sliding (recent weeks near ~1%). This is not a budget decision: IG spend is flat (−2% YoY). Meta is delivering more, cheaper inventory — CPM is down 18%, so the same money buys +19% impressions and +22% clicks. But the extra clicks convert worse: IG purchases are down 17%, and cost per order has risen from $100 to $117 (recent weeks ~$138). In short: more, cheaper traffic — fewer sales.

Important context: IG is the worst-hit, but conversion fell across every source April→June (sitewide −17%) — so about half of this is a site/offer problem, not media (full picture in the company CAC briefing). Below is the IG-media half the team can act on now.

What's driving the volume

It's highly concentrated. One campaign, conv_all_ABO_sandbox, is responsible for 71% of all our Instagram impressions (that's its share of total IG traffic — the campaign also runs on Facebook). It delivers mostly on Reels, now 54% of IG impressions and our lowest-converting placement (2.2% vs Feed's 2.9%). Two ads alone — the "Founder vs Mushroom Coffee" and "Mud vs Mushroom Coffee" videos — make up 54% of all IG impressions.

Why it changed since April

The conversion drop (2.6% → 2.1% of clickers buying) is a mix shift, not one campaign breaking. Three moves pushed the blend toward worse traffic: (1) we cut our best-converting scaled campaign — partner-SHA ASC, ~2.9% CVR — by 64%; (2) we launched ~6 new test/scaling campaigns, several converting under 1.5%, including one ASC scaler at 0.96% ($257/order); (3) a batch of influencer/partner Reels videos is converting at 0.4–1.2% (the worst at $465/order). We replaced proven, high-intent volume with broad Reels, experiments, and influencer content that converts at about half the rate.

The call

1. Pause the influencer/partner leak ads now — 0.4–1.2% CVR, ~400k impressions/mo at 2–4× target cost per order.
2. Kill or fix the 0.96% ASC scaling test (and the coffee-ASC, which fell 3.7% → 1.5%).
3. Restore the partner-SHA ASC scaler we cut — it's our best-converting scaled IG campaign.
4. Cap Reels in the sandbox, or split a Feed-weighted version, and compare cost per order.
5. Manage IG to cost-per-order, not sessions. The traffic growth is vanity — orders are down 17%.
6. Biggest lever sits outside media: conversion fell on every source, so push web/retention to confirm the April→June offer/promo change and any checkout regression since ~May 18. That's what restores blended CAC; the IG fixes above stop the bleeding on the worst channel.

Validated two ways (Meta ad platform + Shopify). One caveat: a site tracking gap on Jun 12–17 understates the last 1–2 weeks; the trend holds regardless.