---
title: "Predicted LTV (pLTV): what it is and when to bid on it | Product Metrics"
description: "Predicted LTV (pLTV) estimates what a customer will be worth. See how it differs from observed LTV, three ways to estimate it and what to check before bidding."
canonical: "https://www.productmetrics.io/blog/predicted-ltv"
pageType: article
language: en
publisher: "Product Metrics"
author: "Berend Vrakking"
datePublished: 2026-10-07
dateModified: 2026-10-08
image: "https://www.productmetrics.io/og/blog/predicted-ltv.png"
---

> Content index: https://www.productmetrics.io/llms.txt

# Predicted LTV (pLTV): what it is and when to bid on it

## Key takeaways

- Predicted LTV (pLTV) estimates what a customer will produce in a set window. Observed LTV is what they have already produced, so only it can be checked against orders.
- A bid is placed on a first order: €40.00 of the €88.00 predicted for 12 months in our example. A prediction values a buyer before repeat orders exist.
- An over-estimate pays too much: a prediction of €88.00 against an observed €70.00 allows €9.00 too much per customer if you want to keep half the profit.
- Check every prediction against observed profit once its window has passed. Google's 'New customer lifetime value' metric is a conversion value adjustment for first purchases, not a forecast.

Predicted LTV, often shortened to pLTV, is an estimate of what a customer will be worth in a set window, such as the next 12 months. "pLTV" on its own also turns up in baseball, so in this guide it always means the ecommerce metric.

It is written for PPC specialists and ecommerce owners who bid for new customers on Google Shopping and want to know how much faith to put in a number that hasn't happened yet.

## How is predicted LTV different from observed LTV?

Observed LTV is the revenue or profit a customer has produced so far, and predicted LTV is an estimate of what they will produce in a window that has not ended. The first is counted from orders, and the second has to be checked against them later.

| | Observed LTV | Predicted LTV (pLTV) |
|---|---|---|
| What it is | What a customer has produced to date, including order revisions | An estimate of what they will produce in a window |
| When it is known | After the orders happen | Right after the first order |
| Can it be checked? | Yes, against orders | Only after the window has passed |
| Main risk | Looks low for new customers | Over-estimates pay too much for them |

Observed LTV includes order revisions: a refunded or cancelled order leaves the figure. It is the number behind [LTV:CAC for ecommerce](https://www.productmetrics.io/blog/ltv-cac-ratio-ecommerce), and it can be counted from orders without any forecast.

Both numbers need a window and a measure. "LTV" alone does not say whether it is revenue or profit, or whether it covers 3, 12 or 24 months, so a prediction is only comparable with an observed figure measured the same way. Profit is the better measure for bidding, because two customers with the same revenue can leave very different profit.

> **Sidenote:** Google's "New customer lifetime value" metric is neither of these. Google Ads Help defines it as "the conversion value adjustment corresponding to acquisition conversions (first purchase conversions determined to be from new customers)". It is a bidding value added to a first purchase, not a forecast of what the customer will spend.

## How do you estimate predicted LTV?

You can estimate LTV from a historical average per customer, from a cohort curve built on past orders, or with a statistical or machine-learning model. The first two are arithmetic you can do in a spreadsheet, and each one breaks in a different place.

| Method | What it does | What it needs | Where it breaks |
|---|---|---|---|
| Historical average per customer | Gives every new customer the average profit of past customers over the same window | Customers who have had the full window | Treats all new customers alike, and fails when the product mix or prices change |
| Cohort curve | Groups customers by first-order month, follows their cumulative orders, and extends that curve | Several cohorts with months of history | Seasonal products, small cohorts and a one-order customer, who gets the cohort's average |
| Statistical or machine-learning model | Uses what is known at the first order, such as the product bought, to estimate each customer | Many customer-level orders, and a check against later orders | Small shops, and any change the history has not seen |

The hardest customer to estimate is the one with a single order, which is every new customer at bid time. Nothing about their behaviour after the first order exists yet, so any method gives them the average of similar customers. Only a model that uses what is known at the first order, such as the product bought, can say more, and it needs plenty of past orders to learn from.

A [cohort](https://www.productmetrics.io/glossary/cohort-analysis) curve is the easiest to explain. If a cohort placed 1.30 orders per customer by month 3, you can extend that pace and say what profit per customer it implies at month 12. Your [repeat purchase rate](https://www.productmetrics.io/glossary/repeat-purchase-rate) is the main input to it.

Small shops are where all three struggle. A handful of large buyers can move an average a long way, which is why an estimate built on 20 customers deserves less trust than one built on 2,000.

**Figure: How far can one big buyer move average orders per customer?.** Each bar is the average number of orders per customer in a cohort. One customer with 10 orders lifts a 20-customer average from 1.30 to 1.75, but a 2,000-customer average by less than 0.01.

| Cohort size, with or without one 10-order buyer | Average orders per customer |
| --- | --- |
| 20 customers | 1.30 |
| 20 customers, one 10-order buyer | 1.75 |
| 2,000 customers, one 10-order buyer | 1.30 |

_Source: Calculated as orders ÷ customers: 26 orders for 20 customers, 35 orders after one customer's 1 order becomes 10, and 2,609 orders for 2,000 customers. Illustrative data_

## One cohort, counted then extended

In practice the observed part of a customer's value comes first and the predicted part is added to it. The chart below follows one cohort: the first months are counted, and the later months are an extension of that pace.

For example, a customer's first order brings €40.00 of profit. By month 3 the cohort has placed 1.30 orders per customer, so profit is €52.00. If the pace holds, orders reach 2.20 per customer by month 12, which is €88.00.

**Figure: One cohort's profit per customer, in €: observed to month 3, predicted after.** The line is the total profit per customer: months 0 and 3 are observed (€40.00, €52.00), months 6 to 12 predicted (€64.00 to €88.00). The dashed line is the €60.00 it cost to win each customer, which the prediction says is repaid between month 3 and month 6.

| Months since first order | Total profit per customer so far (€) |
| --- | --- |
| Month 0 | €40.00 |
| Month 3 | €52.00 |
| Month 6 | €64.00 |
| Month 9 | €76.00 |
| Month 12 | €88.00 |

Reference line: Cost to win each customer (nCAC) (€60.00).

_Source: Calculated as cumulative orders per customer (observed 1.00 and 1.30, predicted 1.60, 1.90 and 2.20) × €40.00 profit per order. Illustrative data_

The dashed line is a target, here an acquisition cost (nCAC) of €60.00. The prediction says the cohort repays it between month 3 and month 6. That statement is a forecast until month 6 arrives. The over-estimate section below shows what happens when it turns out wrong, and [customer acquisition cost for ecommerce](https://www.productmetrics.io/blog/customer-acquisition-cost-formula) explains how to work out nCAC. The [MER and nCAC calculator](https://www.productmetrics.io/mer-calculator) gives your own figure from ad spend and new customers.

## Why bid on a prediction at all

When Smart Bidding places a bid, the customer has not bought again, so the first order is all there is to value. A prediction lets you value that buyer for what they may do next instead of for the basket alone.

**Figure: How much of a customer's 12-month profit, in €, is known at bid time?.** Each bar is one part of the €88.00 profit predicted per customer over 12 months. At bid time only the first order's €40.00 exists, less than half; €12.00 arrives by month 3 and €36.00 is still a prediction.

| Part of the profit, and when it is known | € profit per customer |
| --- | --- |
| First order (known at bid time) | €40.00 |
| Repeat orders to month 3 | €12.00 |
| Predicted, months 3 to 12 | €36.00 |

_Source: Calculated as €40.00 first order, +€12.00 by month 3 (1.30 − 1.00 orders × €40.00), +€36.00 predicted (2.20 − 1.30 orders × €40.00). Illustrative data_

Without a prediction you either bid on the first order's profit alone, or you wait months for observed value. Google Ads Help describes the trade-off: "Short-term conversion values can be useful when you want to maximize immediate profit or customer acquisition as cash flows allow. Lifetime conversion values can be more useful when trying to maximize long term growth."

Google also says that, when you can't track a value definitively, "using a 15-20% conservative estimate is often more helpful than using no value at all." That sentence is about unmeasured gains such as word of mouth, and not about LTV. The principle carries over: a cautious estimate you can defend beats no value, so shade a prediction down before you bid on it.

## What an over-estimate costs

If a prediction is too high, you pay too much for every new customer it covers, and you only find out when the window ends. The cost grows with the gap between predicted and observed profit.

Take the cohort above. At month 3 the prediction says €88.00 by month 12. The cohort then orders less often, and reaches €70.00. The gap is €18.00 per customer.

**Figure: Predicted at month 3 against what happened: profit per customer, in €.** Both lines are total profit per customer; after month 3 the prediction runs ahead and ends at €88.00, €18.00 above the observed €70.00. The dashed line is the €60.00 nCAC: the cohort repays it between month 6 and month 9, not between month 3 and month 6.

| Months since first order | Predicted at month 3 (€ per customer) | Actually observed (€ per customer) |
| --- | --- | --- |
| Month 0 | €40.00 | €40.00 |
| Month 3 | €52.00 | €52.00 |
| Month 6 | €64.00 | €58.00 |
| Month 9 | €76.00 | €64.00 |
| Month 12 | €88.00 | €70.00 |

Reference line: Cost to win each customer (nCAC) (€60.00).

_Source: Calculated as cumulative orders per customer × €40.00 profit per order (predicted 1.00, 1.30, 1.60, 1.90, 2.20; observed 1.00, 1.30, 1.45, 1.60, 1.75). Illustrative data_

Suppose you want to keep half of each customer's profit after paying to win them. Then the most you can pay is half the value you used.

**Figure: To keep half the profit, how much can you pay per customer, in €?.** Each bar is half of the customer profit it is based on: the most you can pay to win a customer and keep the other half. The predicted €88.00 allows €44.00, but the observed €70.00 supported only €35.00, so the over-estimate gave away €9.00.

| Customer profit the limit is based on | Most you can pay per customer (€) |
| --- | --- |
| First order only (€40.00) | €20.00 |
| Predicted 12 months (€88.00) | €44.00 |
| Observed 12 months (€70.00) | €35.00 |

_Source: Calculated as (1 − 50%) × customer profit. Illustrative data_

The first order alone supports €20.00, the prediction €44.00 and the observed result €35.00. Paying on the prediction gave away €9.00 per customer. Checking it takes a few steps, and [LTV:CAC for ecommerce](https://www.productmetrics.io/blog/ltv-cac-ratio-ecommerce) shows how to compare the result with nCAC.

**Check a prediction before and after you bid on it**

1. **Name the window and the measure.** Write down whether the prediction is revenue or profit, net of returns, and for how many months, for example profit over 12 months.
2. **Bid on a cautious version first.** Start with a value you are sure the cohort will reach, since Google says a conservative estimate is often more helpful than no value at all.
3. **Wait for the window to end.** Group customers by first-order month and follow each cohort until the window has passed, so the comparison uses a full period.
4. **Compare with observed profit.** Set observed profit per customer next to the prediction. If the prediction was higher, lower it before the next cohort.
5. **Re-check each cohort.** Seasonal products and price changes shift repeat buying, so a prediction that worked last quarter needs checking again.

## Do Google's lifecycle goals predict LTV?

No. Google's customer lifecycle goals tell Smart Bidding to treat new customers differently. Google Ads Help says they "help you increase value from both new and existing customers", and lists them for Performance Max, Search, Shopping and Demand Gen campaigns.

The acquisition goal New Customer Value is described as "Bid higher for new customers than existing ones". That is the adjustment the metric in the sidenote above reports: a higher value on first purchases from new customers. Whatever its name suggests, it looks no further ahead than the first basket.

Prediction is optional; observed profit is the safer place to start, and it is one of the conversion actions [Product Metrics](https://www.productmetrics.io/server-side-tracking) sends to Google Ads, alongside new customers, returning customers and lifetime value. For the break-even floor any value starts from, use the [break-even ROAS calculator](https://www.productmetrics.io/break-even-roas-calculator), and see [POAS vs ROAS](https://www.productmetrics.io/blog/poas-vs-roas) for the profit version.

## Observed first, prediction second

Start with observed profit per customer by first-order month, so you know what your cohorts produce. Only then decide whether a prediction would add something, and write its window and measure down before you use it.

When the window ends, compare prediction and observed profit, and lower the next prediction if it was too high.

## Frequently asked questions

### What is predicted LTV?

Predicted LTV (pLTV) is an estimate of the revenue or profit a customer will produce over a set window, such as the next 12 months. Unlike observed LTV it can't be counted from orders yet, so it has to be checked against what customers later spend.

### What is the difference between historical and predictive LTV?

Historical LTV, which we call observed LTV, adds up what a customer has ordered so far, including order revisions. Predictive LTV estimates what they will order next. Observed LTV is a fact about the past, and predictive LTV is a forecast you can test once the window has passed.

### How accurate is predicted LTV?

It depends on the data and the window. Accuracy is lower for customers with one order, small shops where a few buyers move the average, seasonal products and longer windows. Check it by comparing the prediction with observed profit per cohort after the window ends.

### Does Google Ads use predicted LTV?

Not as a forecast, going by Google Ads Help. Its pages on customer lifecycle goals and metrics describe bidding higher for new customers, and a 'New customer lifetime value' metric defined as the conversion value adjustment for first purchases from new customers. Nothing on those pages predicts lifetime value.

---

Written by Berend Vrakking, founder of Product Metrics. Last updated 2026-10-08.

HTML version: https://www.productmetrics.io/blog/predicted-ltv