Is It CAC Inflation or LTV Decay? Isolating the Driver Behind a Broken Ratio
A 30-minute diagnostic that isolates whether your sub-1 LTV:CAC ratio is driven by CAC inflation or LTV decay — so your first fix lands on the right team.
Quick answer
Compare the last 90 days against the prior 90 across two numbers only: blended CAC and 90-day cohort revenue per acquired customer. Whichever moved more than 15% is your driver. If CAC moved more, escalate to media buying. If 90-day cohort revenue moved more, escalate to retention or merchandising. If both moved less than 15%, the ratio break is mix-shift, not a real decay — fix attribution before you fix anything else.
CAC inflation vs LTV decay diagnosis
A 30-minute check that isolates whether a sub-1 LTV:CAC ratio is driven by rising acquisition cost or shrinking customer value.
When LTV:CAC drops below 1, the ratio itself tells you nothing about which side broke. CAC inflation and LTV decay produce the same symptom — an unprofitable customer — but demand opposite responses. CAC inflation is a media-buying problem: auction pressure, channel saturation, or creative fatigue. LTV decay is a product or retention problem: weaker repeat rates, smaller baskets, or a worse cohort quality. The diagnostic isolates which numerator or denominator moved, in what magnitude, before any team starts spending cycles on a fix.
The cost of guessing is two weeks. A performance manager who assumes CAC inflation will pause campaigns, rebuild creative, and renegotiate with agencies — while the real problem was a checkout regression killing repeat purchases. Meanwhile the retention team chases a phantom churn signal.
Step 1: Pull the two numbers that actually matter
Open GA4 or your warehouse and pull two figures for the last 90 days and the prior 90 days: blended CAC (total paid spend ÷ new customers) and 90-day cohort revenue per acquired customer.
Use 90-day cohort revenue, not lifetime LTV. Lifetime models are too slow to catch a recent break, and they mix old healthy cohorts with new broken ones — exactly the signal you're trying to isolate.
Don't use trailing LTV here
A 12-month trailing LTV will still look healthy for months after a real cohort break, because eight of those twelve months reflect pre-break behaviour. Use the 90-day cohort window or you'll diagnose the problem after it's already cost you a quarter.
Step 2: Apply the 15% rule
Calculate the percentage change on both metrics. The side that moved more than 15% is your driver. The 15% threshold filters out normal seasonal noise — anything below it is signal-to-noise indistinguishable from a slow Tuesday.
If CAC rose 28% and cohort revenue fell 4%, you have CAC inflation. Escalate to media buying: audit channel mix, check Meta CPM trends against your vertical, and review creative fatigue scores. If CAC moved 6% and cohort revenue dropped 22%, you have LTV decay. Escalate to retention and merchandising: pull the cohort decay curve to see which week the drop started.
Step 3: Handle the ambiguous case
Sometimes both metrics moved less than 15% but the ratio still flipped below 1. That's not a real break — it's mix shift or attribution drift, and fixing CAC or LTV won't help.
Common causes: a new channel with higher CAC scaled fast enough to drag the blended average, or a discount campaign pulled forward revenue that would have landed in the next cohort window. Check channel-level CAC and promo calendar before touching anything else.
Both moved more than 15%? You have two problems.
Rare but real, usually during a category downturn — beauty in Q1 after a holiday push, apparel after a weak collection drop. Address the larger mover first, then revisit the second in 30 days. Trying to fix both at once means neither gets a clean read.
What to do with the answer
Diagnosis isn't the fix — it's the routing decision. CAC inflation routes to a media-buying triage: channel reallocation, creative refresh, bid strategy review. LTV decay routes to a cohort autopsy: which segment, which product, which acquisition source produced the weaker customer. The 14-day triage playbook covers what to do once the driver is known.
If you want to catch the next break before it costs a quarter, the cohort decay leading-indicator method spots sub-1 LTV:CAC about three months before it shows up in blended numbers. Same diagnostic logic, applied weekly instead of after the fact.
Frequently asked questions
Compare percentage change in 90-day blended CAC versus 90-day cohort revenue per acquired customer, over the last 90 days against the prior 90. Whichever moved more than 15% is the driver. If both moved less, it's mix shift or attribution drift, not a real ratio break.
Full LTV includes months of pre-break customer behaviour and lags reality by a quarter or more. A 90-day cohort window isolates recent acquisitions, so a break this quarter is visible this quarter — not next year when the trailing LTV finally rolls over.
Don't pause campaigns or launch a retention sprint yet. Run the two-number diagnostic first: blended CAC change vs 90-day cohort revenue change. The 30 minutes spent here saves the two weeks you'd otherwise burn fixing the wrong side.
Route to media buying. The likely causes are auction pressure (check Meta and Google CPM trends in your vertical), channel saturation (one source taking too much spend share), or creative fatigue. The 14-day triage playbook details the first-lever decision.
Route to retention and merchandising. Pull a cohort decay curve to see which acquisition week first showed the weaker revenue. Common culprits: a checkout regression, a product-quality slip in a hero SKU, or a discount cohort that converted but never came back.
Below 15%, the change is usually within seasonal and channel-mix noise — you can't act on it confidently. Above 15%, the move is large enough to survive a sanity check across weekly slices and warrants a real intervention. Adjust the threshold down to 10% only if your traffic is stable enough that weekly numbers don't swing.
Yes — pull new-customer counts from Shopify orders and paid spend from your ad platforms (or a connector). GA4 is fine for the cohort revenue side as long as you've got purchase events firing correctly. If GA4 attribution is broken, fix that before running the diagnostic, otherwise both numbers are noise.
Use 60-day cohort windows and accept slightly noisier signal. Don't go below 60 days — early repeat purchases skew toward your most enthusiastic customers and overstate cohort revenue. If you're under 60 days of trustworthy data, the answer is to instrument now and wait, not to diagnose on bad numbers.
Yes. The two metrics, the 15% threshold, and the routing logic are all rule-based. Most teams run the check weekly inside their analytics platform with an alert when either side crosses the threshold — which surfaces the break weeks before the blended ratio rolls below 1.
This diagnostic runs after the ratio has already broken — it answers which side to fix. A cohort decay leading indicator runs before the break, watching weekly cohort revenue curves to predict a sub-1 ratio about three months out. Pair them: leading indicator for prevention, this diagnostic for triage.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.