Table des matières
- Introduction : Poser les bases de la gestion de produits axée sur les données
- 1. Taux de rétention des revenus nets (NRR)
- 2. Adoption
- 3. Stickiness
- 4. Growth
- 5. Product Engagement Score (PES)
- 6. Fidélisation
- 7. Délai de rentabilisation
- 8. Net Promoter Score (NPS)
- 9. Fonctionnalités les plus demandées
- 10. Performance du produit
In a rush?
Download the PDF for later
Introduction : Poser les bases de la gestion de produits axée sur les données
Si vous demandez à cinq responsables produit de nommer les indicateurs les plus importants à suivre, vous obtiendrez probablement (voire certainement) cinq réponses complètement différentes, et de quoi susciter de vifs débats. En effet, lorsqu'il s'agit de données, les responsables produit se trouvent aujourd'hui face à un paradoxe de choix. Parmi la multitude d'indicateurs produit à suivre, il n'est pas toujours évident de savoir lesquels privilégier.
À une époque, l'accès aux données produit était difficile. Cette période désormais est révolue. De nos jours, les spécialistes en produits de tous niveaux doivent cibler les indicateurs qui auront le plus d'impact ; ceux qui les aideront à comprendre comment les utilisateurs parcourent leur produit, à évaluer en quoi l'utilisation du produit génère des résultats positifs et à savoir comment le produit joue sur les résultats financiers de l'entreprise.
En adoptant une stratégie de mesure efficace, les responsables produit peuvent baser leurs décisions sur des données et mettre au point des produits qui comblent (et dépassent) les besoins des utilisateurs, favorisant ainsi le succès de l'entreprise. Les entreprises orientées produit vont jusqu'à rallier tout le monde autour des données produit, faisant de celles-ci une ressource centrale et un langage commun. Dans ce contexte, il leur est essentiel de veiller à mesurer les bons indicateurs.
Les ICP qu'une entreprise doit suivre varient en fonction de son stade de croissance, du produit proposé et de ses objectifs globaux. D'après nous, il existe néanmoins dix indicateurs auxquels tout responsable produit doit prêter attention. Que vous soyez pleinement impliqué dans les activités quotidiennes ou davantage intéressé par une vue d'ensemble des opérations, les indicateurs présentés dans ce guide constituent une solide base pour votre stratégie d'analyse produit.
Résultats commerciaux : votre produit a une incidence sur vos résultats commerciaux et financiers, aussi bien à court terme qu'à long terme. L'expérience produit joue notamment un rôle direct sur la fidélisation et l'attrition, ce qui se répercute sur le chiffre d'affaires.
Utilisation du produit : comme le nom l'indique, ces indicateurs reflètent le comportement des utilisateurs au sein de votre produit. Quelles sont les fonctionnalités qu’ils utilisent le plus ? Où restent-ils bloqués ? Combien d'utilisateurs continuent d'accéder au produit au fil du temps ? Si les données relatives à l'utilisation du produit sont utiles en soi, elles sont d'autant plus pertinentes lorsqu'elles sont associées à des mesures qualitatives telles que le sentiment et le retour client.
Qualité du produit : ces ICP évaluent la capacité de votre produit à remplir son rôle. Quel est le temps de réponse de votre produit ? À combien s'élevait le temps d'arrêt le mois dernier ? Pour les indicateurs de qualité, l'idéal est de définir une valeur de référence en interne et d'y comparer vos résultats tous les mois ou trimestres.
1. Taux de rétention des revenus nets (NRR)
À combien s'élève la rétention des revenus générés par les clients existants ?
Résultats commerciaux
Le taux de rétention des revenus nets (NRR) représente le pourcentage de revenus récurrents générés par les clients existants que vous retenez au cours d'une période donnée. Autrement dit, c'est la différence entre l'expansion et la perte de clients. En général, on espère atteindre une valeur supérieure à 100 %, en partant du principe que l'on va perdre quelques clients, mais aussi trouver des façons d'augmenter la valeur des clients qui restent.
Comment calculer le NRR ?
Comme le suggère sa définition, le NRR est calculé en soustrayant la perte de clients à l'expansion. Vous aurez besoin de quatre valeurs :
- Revenus récurrents du mois dernier
- Les revenus provenant de l’expansion (ventes incitatives et ventes croisées)
- Revenue lost due to down-sells
- Revenue lost due to churn
Ensuite, le calcul est le suivant :

En général, les entreprises suivent le NRR sur une base mensuelle ou annuelle. Nous vous conseillons toutefois d'analyser votre taux de rétention des revenus nets chaque mois afin de toujours garder un œil sur les éventuelles fluctuations.
Pourquoi le NRR est-il important ?
NRR is a particularly valuable metric to track because it is all-encompassing—it takes revenue, expansion, and churn all into account. Product teams should always be looking to increase net revenue retention since it represents key goals of expanding customer value and minimizing churn. And there’s no question executives, investors, and the rest of your company will want to know how much revenue your product is able to generate and retain.

2. Adoption
Les utilisateurs adoptent-ils le produit et ses fonctionnalités clés ?
Product usage
L'adoption est une mesure de l'activation qui permet de savoir combien d'utilisateurs interagissent avec votre produit (adoption du produit) ou avec des fonctionnalités données de ce produit (adoption des fonctionnalités). En comprenant comment (ou si) les clients adoptent et utilisent votre produit, vous pourrez savoir clairement s'il leur offre la valeur ajoutée escomptée.
How to measure adoption
As mentioned above, there are two types of adoption: product adoption and feature adoption. Here’s a breakdown of how to measure each:
Adoption globale du produit
L'adoption du produit peut être exprimée au fil du temps par le nombre d'utilisateurs actifs mensuels (MAU), hebdomadaires (WAU) ou quotidiens (DAU). Vous pouvez également mesurer le taux d'adoption du produit en vous basant sur le nombre de nouvelles inscriptions sur une période donnée. La méthode que vous choisirez pour mesurer l'adoption du produit dépendra largement de votre définition du terme « utilisateur actif ». Si votre logiciel est destiné à être utilisé au quotidien, le DAU est sûrement l'indicateur le plus pertinent. En revanche, si vous proposez un produit B2C, vous pourriez analyser la fréquence des conversions (par exemple, achats) ou le temps passé sur l'application.

Adoption de fonctionnalités individuelles
La mesure de l'adoption de fonctionnalités est similaire à celle de l'adoption du produit, sauf qu'elle concerne une ou plusieurs fonctionnalités du produit. L'un des moyens couramment utilisés pour suivre l'adoption de fonctionnalités consiste à évaluer le pourcentage de fonctionnalités qui génèrent 80 % du volume de clics.
Chez Pendo, nous vous conseillons de réaliser cette mesure en tenant compte des fonctionnalités les plus importantes de votre produit, également appelées Actions clés. En vous concentrant sur l'adoption des fonctionnalités liées aux fonctions essentielles de votre produit, vous saurez directement si les clients adoptent les fonctionnalités qui génèrent des résultats positifs. Une fois que vous avez identifié les Actions clés, mesurez l'adoption de fonctionnalités grâce à la formule mathématique suivante :

One key time to measure feature adoption is after you’ve launched a new feature. Use your product analytics tool to look at what percentage of users initially adopted the feature and, more importantly, how many continued to use it after your promotional campaign tapered off. We suggest looking 30 days after a launch to understand drop-off patterns.
Vous obtiendrez également de précieuses informations en analysant l'adoption de fonctionnalités au niveau des utilisateurs et des comptes. L'analyse des résultats au niveau des utilisateurs vous permettra de mieux comprendre le comportement de votre type de client cible, tandis que cette même analyse au niveau des comptes (c'est à dire, par entreprise) vous aidera à écarter ceux qui n'ont pas eu besoin de la fonctionnalité en raison de leur rôle ou de leur niveau d'autorisation.

Why adoption matters
L'avènement du software as a service (SaaS) accroît la pression sur les équipes, qui doivent proposer des expériences produit capables de maintenir à long terme l'engagement des utilisateurs. Sans cela, une fois venu le moment du renouvellement, rien ne les empêchera d'abandonner votre logiciel et de trouver une alternative. L'adoption du produit et des fonctionnalités sont de précieux indicateurs pour savoir si les utilisateurs découvrent la valeur d'un produit. Ces deux indicateurs sont liés à la fidélisation et à l'expansion. En substance, l'adoption est le signe d'un logiciel et d'une entreprise prospères.
After launching a new feature, MineralTree realized the feature’s placement in their application was making it difficult for users to find. In order to improve adoption, they leveraged an in-app guide to “reposition” the feature on their dashboard page—helping users easily locate the new functionality from the page they visit most often. Since launching the in-app guide, MineralTree has seen a 75% to 100% increase in traffic to the new feature. See how they did it.
3. Stickiness
Do users keep coming back to the product?
Product usage
As a product leader, you need to build software that not only attracts new users, but ensures those users stick around and continuously engage with the product. One way to understand if you’re doing this is through stickiness, which measures how many users return to the product on a regular basis.
How to measure stickiness
La meilleure façon de mesurer la régularité consiste à calculer un ratio. Trois méthodes s'offrent à vous :
- Monthly users who return daily (DAU/MAU)
- Weekly users who return daily (DAU/WAU)
- Utilisateurs mensuels qui reviennent chaque semaine (WAU/MAU)
To determine which ratio to use, start by identifying what ideal engagement with your product looks like. Some products (e.g. a project management tool) are meant to be used daily, while others (e.g. a healthcare patient portal) are geared toward weekly or even monthly usage. Once you’ve determined your product’s ideal engagement levels, you can select the stickiness measure that aligns best with your needs.
Pourquoi la régularité est-elle importante ?
À l'exception des rares applications logicielles qui ne sont destinées à être utilisées qu'une (ou quelques) fois, tout responsable produit souhaite que les utilisateurs continuent à s'intéresser à son produit et à y trouver de la valeur. La régularité vous aide à évaluer la fréquence des visites des utilisateurs, mais aussi à récolter de précieux insights qui vous aideront à booster l'engagement de votre clientèle. Par exemple, après avoir examiné les comportements de vos utilisateurs les plus réguliers, encouragez les autres utilisateurs à suivre des parcours similaires (via des guides in-app et des tutoriels, par exemple).

4. Growth
Is the product acquiring and retaining new users faster than existing users are abandoning it?
Product usage
Growth measures the net effect of your user acquisition and retention efforts. Whether it’s achieved by adding new accounts or increasing usage within existing accounts (or ideally, both), growth is a common top-line metric for modern product teams—and a key indicator of product engagement.
Comment calculer la croissance ?
L'une des façons les plus simples de mesurer la croissance produit consiste à suivre le taux de croissance des utilisateurs ou des comptes sur une période donnée. Chez Pendo, nous utilisons la méthode du Quick Ratio, car ce modèle permet de mesurer à la fois la croissance, la fidélisation et l'attrition. En somme, le Quick Ratio indique l'efficacité avec laquelle votre produit se développe. Il permet de prendre le pouls de votre produit, puisqu'il représente le nombre d'utilisateurs qui l'adoptent, y restent et le quittent sur une période donnée.
Pour calculer la croissance sur une période donnée, utilisez la formule mathématique suivante :

Pourquoi la croissance est-elle importante ?
While it’s easy to get caught up in making the product better, faster, and easier to use, teams still must ground themselves in an engagement metric like growth. If you’re not able to effectively grow your user base—and keep the users you already have—the work you and your team put into improving your product becomes a lot less relevant.

Top Hat’s marketing organization was tasked with helping their peers in customer success with a large renewal campaign. After seeing limited success with email-driven outreach, the team turned to their product itself as a marketing channel, creating in-app guides to automate the renewal campaign and reduce administrative work for their account reps. Learn more about their product-led growth strategy.
5. Product Engagement Score (PES)
How are users engaging with the product overall?
Product usage
Just as there are many KPIs to measure product success, there are multiple ways to measure product engagement specifically (three of which we covered in the previous sections). The Product Engagement Score (PES) exists to provide product leaders with a single metric to measure how users or customers engage with the product. PES is a composite metric, taking the average of one’s adoption, stickiness, and growth rates. The goal? To make it easy to quickly understand how your product is performing.
Here is a breakdown of the components of the Product Engagement Score:

How to measure PES
Once you’ve calculated values for adoption, stickiness, and growth, you can calculate your PES by simply taking the average of the three numbers and multiplying by 100:

You can choose to measure your Product Engagement Score based on visitors or accounts. If your product is only used by individuals (and not teams), then you’ll likely only want to measure each PES component at the visitor level. If your company is focused on new logo acquisition, measuring growth (and overall PES) at the account level will best reflect those efforts. When in doubt, try to always align your PES configuration with your organization’s current overarching priorities.

Prior to leveraging the Product Engagement Score, IHS Markit’s UX team had to patch together many different metrics to paint a picture of product health. Now, with PES, the team has a single metric they can use to understand product success and friction. PES also provides a common language for teams across the organization to collaborate around things like improving adoption or optimizing onboarding. Here’s a look into their journey with PES.
Why PES matters
One of the biggest benefits of leveraging the Product Engagement Score is that it boils engagement down to one number—giving PMs a single go-to metric for reporting product success. Yet PES is equally useful in helping you understand where you can improve. Since the score takes the average of a product’s adoption, stickiness, and growth values, you can quickly identify which metric (or metrics) is bringing down the average. By digging into each number, examining trends over time, and comparing scores for visitors and accounts, you’ll be able to understand the “why” behind each metric, identify areas for improvement, and experiment with different tactics to boost each component of PES.
Using our own business as a case study, Pendo’s data science team wanted to see if, using only PES metrics, we could predict whether a customer would churn, renew flat, or grow its contract.
The team found that PES is in fact correlated with customer retention. In the months leading up to a renewal, accounts with the highest PES were most likely to renew and expand, companies with slightly lower PES were likely to renew flat, and those with the lowest PES correlated with churn. The separation in scores became apparent around six months before renewal time, which means companies can use PES as a leading indicator for renewals.
6. Fidélisation
Are users building enduring habits inside the product?
Product usage
Retention measures the percentage of users (or customer accounts) still using your product after they initially install or start using it. The two main types of retention are user retention and customer retention: user retention looks at the individual who logs in to use the product, and customer retention looks at the account that pays for access to the product. Another way to think about retention is as the inverse of churn.
How to measure retention
As with measuring adoption, you can measure retention for the product overall, or for specific features. Here’s how to go about each method:
App retention
Since product usage data is most valuable when measured over time, app retention measures how many users or customers continue using an application during a given time period, for example the first month of usage, three months of usage, or six months of usage.
Feature retention
Measuring feature retention gives you an understanding of which features keep your users returning and helps you identify any users who are at-risk. The fewer features a user interacts with (and the less often they interact with them), the less likely they are to continue seeing value in your product. If you’ve established Core Events, it might make sense to focus your retention measurement around these key areas of the product.
To take measurement a step further, we also recommend segmenting your data (e.g. by role, company size, or free users vs. paying customers) to compare feature retention across different subsets of users and determine how their behaviors differ.

Why retention matters
If you’re not able to retain customers, it directly impacts your company’s revenue numbers. Since the cost to acquire a customer can often be greater than the initial contract value, low retention can cause an organization to actually lose money when they sign a new customer. It makes the stakes that much higher for delivering a digital experience that keeps users engaged and active in the product over time.
For Talegent’s sales and customer success teams, product data in Pendo offers an important lens into customer health and engagement. Account managers can flag customers with low engagement, start exploring potential causes, and work to solve any issues before the possibility of churn comes into the picture. See how else they’re boosting retention.
7. Délai de rentabilisation
How long does it take for users to find value in the product?
Product usage
Time to value measures the amount of time from when a customer starts using your product to when they start deriving value from it. Many refer to the latter point as the “aha” moment—when a user has a clear grasp of why they need a specific piece of software.
How to measure time to value
At Pendo, we like to anchor time to value measurement around Core Events, which are the ten (or fewer) key features in a product. In this case, time to value represents how long it takes new users to engage with a Core Event. Ideally, your product analytics tool can tell you the average, minimum, and maximum time it takes new visitors to first interact with a Core Event. Bonus points if you’re able to compare this to the previous time period to see trending changes.
Why time to value matters
No matter how many new features or shiny updates you have on the roadmap, if users don’t feel the benefits of your product right away, they likely won’t stick around long enough to experience those improvements. This puts an added emphasis on a new user’s first impression of your product—they need to understand both how your product works and the value it will bring them.

The UserTesting team used product usage data to understand where users were dropping out of their onboarding flow, and identified a key step that 99% of users were failing to reach (drafting a test). Since they knew engaging with this feature correlated with long-term usage, the team created an in-app onboarding experience that walked users through how to create a test. This led to a 29% increase in users who make it to the “Draft Test” stage and successfully activate the product. Here’s how they did it.
8. Net Promoter Score (NPS)
Are users and customers happy with the product?
Résultats commerciaux
Le Net Promoter Score (NPS) est l'un des moyens privilégiés par les entreprises pour évaluer la fidélité de leurs clients. Il s'agit d'un sondage à question unique visant à déterminer dans quelle mesure vous recommanderiez cette marque à un ami ou un collègue. Les clients répondent sur une échelle de 0 à 10. Bien que certains jugent le NPS trop simpliste, il reste l'un des ICP les plus répandus en matière de satisfaction client.
How to measure NPS
After conducting an NPS survey, mark scores of 9 or 10 as “Promoters,” 7 or 8 as “Passives,” and 0 through 6 as “Detractors.” Then, subtract the percentage of Detractors from the percentage of Promoters to produce your Net Promoter Score.
Here is the formula expressed mathematically:

NPS scores vary greatly by industry, so it’s important to benchmark yourself against those companies that are most similar to yours (if you’re able to find the data). More important is to benchmark yourself against yourself—track NPS over time and look for any changes, especially as they coincide with business events like product updates and marketing campaigns.
Remember to measure NPS at the account and visitor level. Account-level scores will help you understand your average sentiment, whereas user NPS helps uncover if there are any key personas driving your score up or down. Be sure to also include a section in your survey where users can provide written comments, as this information adds valuable context to numerical scores.

Why NPS matters
Net Promoter Score has implications beyond current customer sentiment. Customers’ propensity to advocate for your brand serves as a proxy for customer happiness and, eventually, business growth. As you measure NPS you’ll likely be focused on improving your score in the short term, but it’s equally important to remember the long-term implications of a strong NPS.

After identifying a common theme about their auto-logout security feature, Patientco’s marketing manager took this feedback from an NPS survey and got to work. First, she deployed an in-app guide to users who reported displeasure with the feature to let them know their feedback was heard. Then, she took the feedback to the product and development teams and came up with a technical solution: adding a notification that warned users they were about to be logged out. The result? A 16-point increase in NPS over a six month period. Learn more about their NPS strategy.
9. Fonctionnalités les plus demandées
What do users want from the product?
Résultats commerciaux
The idea of knowing which features your users are requesting is pretty straightforward, but the tricky part is figuring out which requests are the most important. In this case, “top” can mean features that are requested most often, or requests that are tied to customer accounts with the highest ARR, and therefore have the highest business impact potential.
How to measure top feature requests
As a product leader, you’re probably familiar with feature requests coming at you from all angles—one-off conversations with customers, your own sales team, user interviews, in-app surveys, social media, support tickets. The list goes on.
The biggest challenge is making sense of all these requests so you can identify common themes and determine your users’ most pressing needs. This first requires you to establish a centralized place where all feedback lives, ideally in a system that is easy to update and organize and that everyone at your company can access. From there, you can view feature requests as a whole to understand which requests (or types of requests) appear most often, and better prioritize your team’s work.
It’s also useful to analyze feature requests at the user and account level, and segment by things like company size, ARR, industry, NPS response, or subscription type (if you have a free and paid version of your product). The way you choose to segment and analyze requests will depend on your business’ goals. For example if you’re trying to go after a new persona, you’ll want to look at requests from those users specifically to see if there are any patterns. To help prevent churn from your biggest customers, it’s useful to analyze top requests by account to see if there are improvements worth investing in.

Okta’s customer first programs team wanted to ensure there was a single channel for input and output of customer feedback, and a clear way to prioritize requests. They leveraged Pendo Feedback as their tool of choice, allowing users to submit requests directly and the Okta team to easily prioritize based on popularity, revenue impact, and other key parameters. Since implementation, customers have submitted 440 new ideas—a 57% quarter over quarter increase. See their story.
Why feature requests matter
Although this guide has mainly focused on using quantitative data to make better product decisions, qualitative data is equally valuable. Customer feedback and feature requests provide important context—product usage data tells you what users are doing, but feature requests help explain why they’re doing it. And the best product decisions are informed by both.
Still, you can’t just listen to the loudest customer. Teams need a way to prioritize all their incoming requests and view them holistically to understand themes and potential gaps in the product. Bringing the voice of the customer into your planning process will only make your product better, especially when combined with the product usage data you’re already collecting.
10. Performance du produit
Is the product performing efficiently and without bugs?
Product quality
Whether at work or in their personal life, software users expect products to work, and do so at optimal speeds. For the purposes of this guide, product performance refers to the speed at which it can return requests, downtime vs. uptime, and the number of bugs. Taken together, these measures will help your team stay focused on not only building a product that delights users, but one that doesn’t let them down.
How to measure product performance
One of the key ways to measure product performance is through product response time (which measures the time taken by the product’s system to respond to any request). In order to instill this as a priority for your team, we recommend setting an internal performance benchmark and measuring yourself against it monthly or quarterly. For example, you might set the standard as requests delivered in five seconds or less, and hold yourself accountable to maintaining that standard for a sizable portion of your customer base or a sizable percentage of all requests.
You should also monitor your downtime and uptime metrics on a weekly, monthly, or quarterly basis (or a combination of the three).
When it comes to product bugs, there are a few measurement tactics to keep in mind:
- Track the breakdown of product bugs by feature
- Map bugs to product usage and prioritize fixing those that are in the most heavily used areas of your product
- Measure the number of bugs reported vs. the number of bugs you’ve fixed, and track it over time to evaluate how well you’re maintaining your product’s quality and efficiency

Why product performance matters
There’s no denying that users perceive those products that operate quickly and with as few glitches as possible to be the best. While it might be more interesting to track data points like stickiness or retention, you can’t ignore product performance, or assume everything is fine (cue the meme). Product leaders need to stay on top of these metrics, and ensure their entire team understands the importance, too.
After all, a working product is table stakes and an outage or consistent, pesky bug can severely impact your customers’ experience and their perception of your business as a whole. Not to mention it’s never been easier to switch software providers. Especially in SaaS, users are always just one renewal cycle away from going with a competitor who can deliver a similar experience, but faster.
As a product leader, the last time you heard (or said) something along the lines of, “What does the data say?” was probably in the last day—or hour. While operating with a data-first mindset is valuable, it’s even more valuable when you know you’re measuring the right things. We hope this guide helped shed light on the essential metrics that every product team needs to keep top of mind. These ten KPIs will help you unlock the fun part: using data to make better decisions and build a better product for your users.