The Google Analytics Lifetime Value Report: A 2026 Guide

TL;DR

The Google Analytics lifetime value report most people remember was a Universal Analytics feature sitting under Audience, and it no longer exists. What replaced it is the GA4 user lifetime exploration, which is better in one important way and worse in three. Better, because it adds machine learning predictions for purchase probability, churn probability, and predicted revenue. Worse, because it only covers users active since August 2020, it counts every visitor rather than only customers, and the underlying event data disappears after 14 months unless you export it. That last limit is the one that quietly ruins most analysis, since a meaningful customer lifetime value calculation usually needs 18 to 36 months of history. Set retention to the maximum today, turn on the BigQuery export, and treat the numbers in the interface as directional rather than final.

Key Takeaways

• The Google Analytics lifetime value report you remember belonged to Universal Analytics and is gone. Its replacement is the User Lifetime template in GA4 under Explore, and it works differently enough that old guides will mislead you.
• GA4 calculates LTV as the sum of purchase, in_app_purchase, and app_store_subscription_convert events, minus refunds. If those events are not firing with value and currency parameters, the report returns nothing.
• GA4 user LTV covers every user, not only buyers. That makes it much lower than a true customer lifetime value figure and not directly comparable to what your finance team reports.
• GA4 data retention expires exploration data after 14 months at the maximum setting, and the default is two months. The change is not retroactive, so anything already expired is gone.
• Predictive metrics need at least 1,000 returning users who triggered the condition and 1,000 who did not, inside a seven day window across the preceding 28 days. Most properties never qualify.
• Median B2B SaaS lifetime value against acquisition cost sits at 3.2:1 across 939 companies, with 3:1 to 5:1 the healthy band. Above 5:1 usually means underinvestment in growth rather than excellence.
• BigQuery export is free to enable, unsampled, and not subject to the 14 month limit. It also has no historical backfill, which is why turning it on late is expensive and turning it on early costs nothing.

Introduction

If you came here looking for the report under Audience, Lifetime Value, stop looking. That was Universal Analytics, and Universal Analytics no longer collects or serves data. The version of this post that used to sit at this URL walked through that interface, which is why it stopped being useful and eventually stopped ranking. Screenshots of a dead product are worse than no post at all, because they send people hunting for a menu item that will never load again.

What makes this frustrating is that the concept did not go away. Lifetime value is arguably more important now than it was in 2021, because acquisition costs have risen sharply and the gap between profitable and unprofitable growth has widened. B2B SaaS paid acquisition cost climbed 47% in three years. Knowing which channel brings customers who stick around is no longer a nice quarterly exercise. It is the thing that decides whether your budget survives the next planning cycle.

So this rewrite covers what actually exists in 2026. The modern Google Analytics lifetime value report is the GA4 user lifetime exploration, and what follows is how its numbers are computed, the four limits that catch people out, and how to get from a report on a screen to a decision about where money goes. If you are still deciding between platforms entirely, our comparison of Adobe Analytics and Google Analytics covers that separate question.

FAST FACT: The sampling limit for the user lifetime technique is one million users on free Google Analytics and ten million on the paid product. Beyond that, results are sampled and upscaled. (Source: Google Analytics Help)

Where did the Google Analytics lifetime value report actually go?

It moved, changed shape, and lost its old name. In GA4 you find it under Explore, then the template gallery, then User Lifetime. It is an exploration rather than a standard report, which matters more than it sounds, because explorations run on user level and event level data with an expiry date attached. Standard reports run on pre aggregated tables that persist.

Three things changed in the move. The report is now built rather than opened, so you drag dimensions and metrics into rows and values yourself. The default template gives you a generic starting point that almost nobody should keep. And the end date is permanently fixed to yesterday, so you cannot analyse a closed historical window the way you could in the old interface.

The addition that justifies the disruption is prediction. The old report only told you what had already happened. GA4 layers machine learning on top to estimate what is about to happen, which changes the report from a scorecard into something you can act on. Whether you get access to that layer is a different question, and one covered further down.

How does GA4 calculate lifetime value, and what does it leave out?

LTV in GA4 is the sum of a user's revenue generating events since they were first seen, minus refunds. The qualifying events are purchase, in_app_purchase, and app_store_subscription_convert. If those events are not firing, or are firing without value and currency parameters, the exploration returns an empty table and people assume the feature is broken. It is not broken. It has nothing to add up.

FAST FACT: GA4 user lifetime value covers all users of a site or app, not only paying customers, which makes it substantially lower than customer LTV as finance teams define it. (Source: Littledata, 2026)

That distinction causes more confusion than any other part of this report. Your finance team's customer lifetime value counts revenue per customer. GA4's user LTV divides across everybody who showed up, including the large majority who never spend anything. Both numbers are legitimate. They answer different questions, and putting them on the same slide without labelling them is how meetings go sideways.

There are two further gaps worth knowing before you present anything:

• User lifetime data only exists for users who were active on or after 15 August 2020. Someone who bought from you in 2019 and never came back contributes nothing to the report.
• Without a User ID implementation, GA4 falls back to the device based Client ID, which lives in a browser cookie. Cookie expiry, cross device behaviour, and privacy restrictions all fragment a single person into several apparent users, which pushes measured LTV down.

If you sell to logged in customers, wiring up User ID is the single highest return fix on this list. Everything else in the report inherits the quality of your identity layer. Our notes on customer segmentation techniques go deeper on building segments that survive that fragmentation.

How do you build the GA4 user lifetime exploration properly?

The default template is not the report you want. Build your own version of the Google Analytics lifetime value report in about five minutes:

• Open Explore, choose the User Lifetime template, then clear the pre populated rows and values.
• Add First user source and First user medium to Rows. This is what turns the report into a channel quality comparison rather than a vanity total.
• Add LTV: Average and LTV: Total to Values, plus Total users so you can see the denominator behind each average.
• Add a segment that filters to purchasers only. Without it, the percentile metrics are almost always zero, because most users never buy and even the 90th percentile sits at nothing.
• Add Average lifetime engagement duration and First purchase date if you want to see how long the gap is between first visit and first order.

The percentile view is the part most people skip and the part that changes decisions. Each metric can display as a total, an average, or the 10th, 50th, 80th, and 90th percentile. A channel with a $60 average LTV made of many small buyers is a different business than one with a $60 average driven by a handful of large orders, and only the percentile spread tells you which you have. Averages hide that completely, which is worth remembering when someone asks you to summarise the attribution report in a single number.

What are predictive metrics, and does your property even qualify?

GA4 offers three: purchase probability, which estimates the chance a user buys in the next seven days; churn probability, the chance an active user does not return in the next seven days; and predicted revenue, the revenue expected from a user across the next 28 days.

Those metrics are the strongest argument for the new exploration over anything the old report could do, and they feed directly into audience building for Google Ads. Build an audience of users above the 80th percentile for churn probability, and you have a retention campaign that targets people before they leave rather than after.

FAST FACT: To qualify for predictive metrics, a property needs at least 1,000 returning users who triggered the relevant condition and 1,000 who did not, within a seven day period over the preceding 28 days. Predictions stop updating if model quality falls below threshold. (Source: Google Analytics Help)

Read that threshold carefully, because it disqualifies a lot of properties. It is not 1,000 users. It is 1,000 positive and 1,000 negative examples among returning users, refreshed continuously. Low volume B2B sites almost never clear it, and sites that do clear it can silently drop below the line during a quiet quarter, at which point the predictions stop updating without any obvious alert. Check whether the column is still populating before you build a campaign on it.

If you do not qualify, you are not stuck. Cohort based modelling on exported data works at any volume, and statistical approaches such as BG/NBD combined with Gamma Gamma produce per customer forecasts without Google's threshold. That is more work than clicking a template, but it is also more transparent, and you own the model. Our guide to predictive analytics for customer lifetime value walks through that approach.

Why does GA4 data retention break lifetime value analysis?

This is the limit that catches almost everyone, and it deserves more alarm than it usually gets.

FAST FACT: GA4 retains user level and event level data for two months by default and 14 months at maximum. GA4 360 extends to 50 months and starts around $50,000 a year. Changing the setting is not retroactive. (Source: Analyse, 2026)

GA4 data retention only affects explorations, not standard reports, which is the detail most articles bury. Standard reports run on aggregated tables and persist. But the user lifetime exploration is an exploration, so it lives entirely inside the retention window.

Now consider what a real customer lifetime value calculation needs. Cohort LTV curves typically keep climbing for 18 to 36 months before they flatten. Year over year comparison in explorations needs 24 months to be meaningful. At 14 months you get a two month overlap at best. On the two month default, which is what your property is set to if nobody ever changed it, you cannot do cohort work at all.

Two actions follow, and the first takes sixty seconds. Go to Admin, Data Settings, Data Retention, and set event data retention to 14 months on every property you own. It protects the future and does nothing for the past, so the sooner it happens the less you lose.

The second is exporting to BigQuery. It is free to enable on standard properties, captures every raw event without sampling, and is not governed by the retention setting at all. The catch is that there is no historical backfill: the export only captures data collected after you switch it on. Standard export also caps at one million events a day, and a traffic spike past that threshold pauses the export and leaves a permanent hole in your history. Neither catch is a reason to skip it. Both are reasons to do it now rather than next quarter.

How do you turn the report into an LTV to CAC decision?

A number on a screen is not a decision. The Google Analytics lifetime value report earns its place only when it changes which channels get budget next month, and that requires putting lifetime value next to what you paid to acquire it.

Pull average LTV by first user source and medium from the exploration. Pull spend by the same channel from your ad platforms. Divide. Then apply gross margin, because revenue based LTV overstates the real figure by roughly 30% compared with a margin adjusted calculation, and a ratio built on revenue will tell you to spend money you do not have.

FAST FACT: Median B2B SaaS LTV to CAC sits at 3.2:1 across 939 companies, with 3:1 the accepted floor and 4:1 to 6:1 the top quartile. DTC ecommerce typically runs 1.5:1 to 3:1. (Source: Optifai, 2026)

Two rules of thumb are worth internalising. Below 3:1 you are usually paying too much for what a customer is worth. Above 5:1 you are usually underinvesting, because a competitor will happily buy the customers you are declining to bid on. That second one surprises people who treat a high ratio as a trophy.

Pair the ratio with payback period, because they measure different risks. The ratio measures whether the unit economics work. Payback measures whether you run out of cash before they do. Median B2B SaaS payback stretched to roughly 15 to 18 months in 2026, up from 14 in 2023, and a 5:1 ratio with a 24 month payback is a worse business than 3:1 with an eight month payback. Feed both numbers back into paid advertising optimisation rather than optimising campaigns on last click conversion volume, and the data visualization dashboards your team reviews weekly should show the ratio next to the spend that produced it.

What goes wrong in most customer lifetime value calculations?

Five mistakes account for most of the bad numbers presented in planning meetings.

• Comparing GA4 user LTV to finance's customer LTV. They use different denominators. One counts everybody, the other counts buyers. Label which one is on the slide.
• Skipping the margin adjustment. Revenue LTV against fully loaded CAC is a comparison of two different things, and it flatters the answer by about a third.
• Reporting blended LTV as the headline. Surviving cohort value, paired with retention and payback, is a far better health signal than one averaged figure across every customer you ever acquired.
• Trusting predictions that stopped updating. If volume drops below the qualifying threshold, the model quietly stops refreshing. Nothing turns red.
• Chasing CAC reduction when retention is the real lever. A one percentage point reduction in churn raises LTV by roughly 10 to 15%, which is usually a larger and more durable move than shaving acquisition cost.

That last point is worth sitting with. Most teams spend their optimisation effort on the acquisition side because it is easier to measure and faster to change. The compounding value is on the retention side, which is also where artificial intelligence across the customer lifecycle tends to pay off, and where a disciplined experimentation program produces gains that stack rather than reset.

Summary

The Google Analytics lifetime value report you remember belonged to Universal Analytics and is gone. Its replacement, the User Lifetime exploration, sums purchase, in_app_purchase, and app_store_subscription_convert events minus refunds, across every user rather than only buyers, for anyone active since August 2020. Build it yourself with first user source and medium in the rows, a purchaser segment applied, and the percentile columns switched on, because averages hide the distribution that actually changes decisions.

Two constraints decide whether any of it is usable. Predictive metrics need 1,000 positive and 1,000 negative examples among returning users, which most properties never reach. And exploration data expires at 14 months, well short of the 18 to 36 months a cohort curve needs to flatten. Set retention to maximum today and enable the BigQuery export, since neither is retroactive. Then put lifetime value next to acquisition cost with gross margin applied, aim for 3:1 to 5:1, and read the ratio alongside payback period rather than on its own.

Next Steps

If you only do one thing after reading this, change the retention setting. It takes a minute and every day you wait is a day of data you cannot get back. After that, enable the BigQuery export and start building the history a real cohort analysis needs. For the wider reporting picture, see our guide to machine learning and business intelligence in GA4, our notes on retroactive data analysis in Google Analytics for what can be recovered after the fact, and the broader case for using data analytics for business success.

Not sure whether your LTV numbers are telling you the truth? Book a measurement audit with the Lucrative AI team.

Frequently Asked Questions

What replaced the Lifetime Value report from Universal Analytics?

The User Lifetime exploration in GA4, found under Explore in the template gallery. It is the modern Google Analytics lifetime value report, but it is not a like for like replacement. You build it yourself rather than opening a fixed report, it only includes users active since August 2020, and it lives inside the data retention window rather than persisting indefinitely. In exchange you get machine learning predictions the old report never had.

Why is my LTV showing zero in GA4?

Almost always because revenue events are not reaching GA4 correctly. Check that purchase events fire with both value and currency parameters, since GA4 cannot compute lifetime value without them. If totals look right but the percentile columns are zero, that is expected behaviour rather than a bug: most users never buy, so even the 90th percentile can be nothing. Apply a purchaser segment and the distribution appears.

What is the difference between GA4 user LTV and customer lifetime value?

GA4 user LTV averages revenue across every user of your site or app, including the large majority who never spend. Customer lifetime value, as finance and most business literature define it, counts only paying customers and usually applies gross margin. The GA4 figure will therefore look much lower and should never be presented as if the two are interchangeable. Use the GA4 number to compare acquisition channels against each other, not to set a company wide LTV target.

How far back does the GA4 user lifetime report go?

Two limits apply at once. Data only exists for users active on or after 15 August 2020, and event level data expires after 14 months at the maximum retention setting, or two months on the default. The end date is also fixed to yesterday and cannot be changed. If you need a longer window, exporting to BigQuery is the only route, and it only captures data from the moment you enable it.

Do I need BigQuery to do proper LTV analysis?

For anything beyond a rough channel comparison, yes. Cohort curves need 18 to 36 months to flatten and GA4 gives you 14 at best. The export is free to enable on standard properties, ships raw unsampled events, and lets you join analytics data to CRM records so you can measure real customers rather than sessions. The costs are Google Cloud storage and query fees, which stay small for most sites, plus the SQL skills to work with a nested schema.

What is a good LTV to CAC ratio in 2026?

Three to one is the widely accepted floor and 3:1 to 5:1 is the healthy band, with median B2B SaaS at 3.2:1. Top quartile companies run 4:1 to 6:1. Below 3:1 your acquisition spend is eroding margin. Above 5:1 you are probably underinvesting in growth. Always read the ratio alongside the payback period, because a strong ratio with a two year payback is a cash flow problem wearing a good disguise.

Can Google Analytics predict lifetime value directly?

Not lifetime value as such. GA4 predicts purchase probability and churn probability over the next seven days and revenue over the next 28 days, which is a short horizon rather than a lifetime projection. Those metrics are most useful for building audiences and triggering retention campaigns. For genuine long horizon forecasting you need cohort modelling or a probabilistic model built on exported data.