eCPM vs ARPDAU vs LTV: Which Metric Should App Publishers Actually Optimise For?
Your eCPM went up 20% last month. Revenue is flat. How?
This is not a reporting error. It is what happens when a publisher optimises for a metric without fully understanding what it measures and, critically, what it does not. eCPM, ARPDAU, and LTV are not three ways of measuring the same thing. They measure three different dimensions of monetization health and conflating them produces decisions that look correct on one axis while damaging another.
What Each Metric Actually Measures
eCPM: the price of your inventory right now
eCPM is revenue per 1,000 impressions. It is a demand-side signal: how much advertisers are willing to pay for your inventory at this moment, in current market conditions. It contains no information about whether your impression base is healthy, whether your ad load is sustainable, or whether total revenue is growing. A 20% eCPM increase with a 25% drop in impressions produces less revenue than before, and both numbers can be technically accurate simultaneously.
ARPDAU: the daily health signal
ARPDAU is total revenue divided by daily active users. It captures both ad and IAP revenue normalized against the active user base that generated it. Because it moves when monetization changes and when the user base changes, it surfaces trade-offs that eCPM misses entirely. An ad load increase that raises ARPDAU short-term while eroding retention will show the damage in ARPDAU before it shows anywhere else.
LTV: the full cost of every monetization decision
LTV is the total revenue a user generates from first session through churn. It is the only metric that captures the downstream cost of a bad monetization decision. A 30-day ARPDAU increase from aggressive ad load looks like a win in this month's reporting and shows up as a damaged cohort in next quarter's retention analysis, by which point attributing it to the original decision is difficult.
The eCPM Trap
A publisher increases ad frequency. Impressions rise. Revenue rises. eCPM may also rise. All three numbers move positively on a standard dashboard. What does not appear: session time is dropping because ads are firing during flow states. Six to eight weeks later, DAU begins declining as retention erodes. ARPDAU falls even as eCPM holds, because a smaller user base is generating the impressions. Revenue crosses back below the original baseline.
eCPM is the right metric for benchmarking demand quality, evaluating floor price changes, and comparing demand partner performance. It is the wrong primary optimization target for overall monetization strategy because it contains no information about the user base condition underlying it.
ARPDAU: Track It by Cohort, Not in Aggregate
Aggregate ARPDAU across your entire user base conceals more than it reveals. Day-one ARPDAU reflects new users in onboarding. Day-30 ARPDAU reflects retained users running your full ad experience. Track both separately and watch how they move over time.
A healthy trajectory: ARPDAU grows as the user cohort ages. Retained users generate more revenue over time. The pattern that requires immediate attention: ARPDAU stable or rising while DAU is falling. Your monetization is concentrating on a shrinking base. The absolute revenue number may look acceptable today. The trajectory will not.
LTV: Start Tracking It Earlier Than You Think You Need To
LTV typically lives on the UA team's dashboard as a campaign efficiency benchmark. Monetization teams often review it quarterly, if at all. That separation is a mistake.
The ad load, format selection, and trigger timing decisions made in the first 30 days of a user's lifecycle have the largest single impact on their eventual LTV. By the time LTV data is statistically mature, those decisions have already been made for hundreds of thousands of users.
A practical starting point: ARPDAU multiplied by average retained session days gives you a directional LTV that can be tracked from early in a cohort's life. Two configurations might produce identical 30-day ARPDAU. If one produces a day-60 retention rate 12 percentage points higher than the other, its LTV is materially better despite looking identical in short-term reporting.
Which Metric to Prioritise at Each Stage
Stage | Primary Metric | What to Watch For |
Early stage (0-6 months) | ARPDAU | Declining ARPDAU with rising DAU signals extraction problem |
Growth stage | LTV by cohort | Diverging LTV between cohorts signals a config issue |
Mature / stable | eCPM + ARPDAU | ARPDAU falling with stable DAU requires immediate audit |
Post config change | ARPDAU delta | Session time drop within 72hrs = timing problem |
The Diagnostic Checklist
- eCPM declining: audit demand quality, floor price logic, and placement viewability
- ARPDAU declining with stable DAU: investigate recent ad config changes and format mix
- ARPDAU declining with falling DAU: the monetization strategy is affecting retention; audit timing and frequency first
- LTV lower in newer cohorts than older ones: a recent config change is damaging early user experience; find what changed and when
The Number That Actually Matters
eCPM tells you what your inventory is worth per impression. ARPDAU tells you what your users are worth today. LTV tells you what your monetization strategy is worth over time. All three are necessary. None is sufficient alone.
If your reporting shows eCPM prominently and LTV is a quarterly agenda item, that is the configuration worth fixing. YieldSolutions works with app publishers to connect short-term monetization signals to long-term revenue health. If you want to review how your current metrics map to actual decisions, our team is available.
