I haven’t met a product leader at a SaaS business who didn’t obsess about customer retention. Customer Success leaders - by definition - also care deeply. But how do product leaders and teams confidently know what user behaviors truly lead to consistently high rates of retention and what they can do to maximize retention potential and minimize retention risk?
This article is not generically about how to use data to make better product decisions. All good product teams look at product data (in some way) and use it to help make prioritization decisions, to better understand user behavior, and to measure product success (in some way). There are many best practices to follow and analytics tools to use- all good. However, what most organizations are doing is insufficient because it’s missing the most important potential impact of product data: improving business outcomes.
The approaches most organizations take to tie product data to business outcomes (described in more detail below) can be likened to having many people in an inflatable raft paddling furiously with their hands. There is useful activity, you are generally moving in the right direction, but there is limited speed, efficiency, and coordination. My past two companies- as well as several others with whom I have shared the idea- have implemented a new approach called Adoption Ladders.
This can be likened to having stakeholders from Product, Marketing, and Customer Success all rowing in sync as part of a crew in a (sleek) racing shell. I’ll explain further below why this is a significant improvement on the traditional approaches.
Is Customer X a good user of our product? Ways companies try to answer this question today
What is the most foundational business outcome in a SaaS business? Customer retention.
What is a critical driver of retention? Healthy adoption of the products the customers pay for.
Customers who use the product fully and as designed tend to see high value from the product and therefore tend to be at the upper end of your company’s retention range.
How do you help customers get “healthy adoption” and how do you determine if a customer is a healthy adopter? Product, Customer Success, and Marketing all coordinate together to ultimately impact the adoption of the product by customers, but each is likely to offer different answers to this question.
Though adoption is driven by a coordinated effort across Product, Customer Success, and Marketing (and others), many organizations struggle to agree on a shared way to measure and benchmark good adoption so adoption can be easily optimized across all of these teams.
There are two ways that organization typically try to define good usage:
- Look at various product usage metrics to identify good adoption
The natural path for many organizations is to try to look at some set of product usage data. The org may decide on one or two key metrics, such as MAU (monthly active users), that is easy to measure and understand. However, this is overly simplistic and will not accurately distinguish good usage (it’s just overall activity) nor capture the nuances of adoption in a multi-module or multi-product companies.
Alternatively, the org might try to more richly capture the nuances of adoption by looking at a number of metrics for each module/product. A proficient PM may look at a dashboard of 5–10 (or more) metrics to understand how customers are using their product.
This rich dashboard of metrics for each product rarely tells a simple and easy story for each customer and does not clarify what is considered “good” in these metrics. Can these be interpreted consistently and easily by the Product Manager, much less other key stakeholders like the customer success manager?
Because of the complexity of using product data to distinguish customers who are good adopters, customer success organizations and marketing often will turn to a more subjective approach: using maturity models.
2. Use “Maturity Models” to categorize customers into “advanced” and “basic” levels
The Maturity Model is a useful way to describe the levels and steps that a customer might go through from “light/basic” use of the product to “advanced/exemplar” (it can also be used to describe the progression in a market category).
For example, a Maturity Model describing a company adopting Slack might look like
The advantage of this “categorization” model is that it is easy to explain to customers and prospects via marketing, and helps describe how the world changes for a customer when they go from “A” to “B”. The challenge is that assigning a customer to one of these categories is very subjective and customers may exhibit characteristics crossing multiple maturity levels (and may be at different levels in the different products or modules they license).
While undeniably useful, both approaches fundamentally come at the question of “is the customer a good adopter of our products?” from different directions. If different departments use different approaches- eg, Product focuses on product data and CS tends to focus on maturity models- you are speaking two languages and may come up with different answers for the same company.
There are clear benefits to both ways to think about adoption. Product adoption metrics are objective and can be tracked over time and reported on in a scalable way that allows customer-facing teams to take action to further customer adoption. Maturity models offer a nuanced and simple way to describe and bucket customers and are often aspirational.
The good news: Adoption Ladders are designed so these two approaches can work together as one. It has the benefits of the simplicity of the “maturity model” approach but with the rigor of a data-driven approach.
What is an Adoption Ladder?
Adoption Ladders allow companies to define and measure the level of adoption and the overall value being seen by customers and teams to align on adoption levels and act on improving customer account retention.
The Adoption Ladder methodology combines three powerful elements:
- an easily understood framework of adoption maturity levels
- objective measurements to determine where customers are in relation to where they could be
- Alignment, communication and operations across all customer-serving teams.
Teams can determine how healthy any given customer’s current product adoption level is, understand the metrics that define the level, and then take the appropriate actions (product enhancements, customer training, etc) that should improve adoption metrics to progress the customer to the next level.
For each product or product module, a customer’s adoption level can be expressed as a level from 0 to 5.
An “L5” is an exemplar customer — using the product at the breadth, depth, and quality that will deliver high value to them from the product.
An “L1” is a customer who is using the product at the minimum level to derive at least some value (but may not be much).
L2 to L5 are different levels between these. L0 is a company that is not using the product at all or using it such a low level that it does not derive meaningful value.
A few product metrics are used to determine the level to which each customer is assigned.
It is important to note that a multi-product company (or ones with multiple major modules) sold to the same buyer will have multiple adoption ladders — one for each major product. Deriving an overall adoption ladder (across the modules) is described below and is meant to provide an overall view of the value the company derives from using all of the company’s products and modules together.
How to create an Adoption Ladder
There are 5 steps to create an adoption ladder for your customers spanning your product portfolio:
1. Identify metrics for each key product or module
The team should start by identifying 1–3 usage metrics for each product or key module. These metrics are selected because (a) they can be measured automatically and (b) can help distinguish “good” and “weak” usage of the product. These measurements are three types: “activity” (volume or frequency of use), “depth” (effective use that follows best practices), or“breadth” (broad use of the product capabilities).
For example, if the product is a tool that allows requesting and giving feedback to peers, the key metrics might be % of all employees who have requested feedback each month and the % of those requests that result in an employee giving feedback (eg, “response rate”).
How do you find these metrics to start with? In general, you can do this top-down and/or bottom-up. Top-down: best if the product team already has a strong idea of what should differentiate good/bad usage of the product but may or may not already have good access to this data. Bottom-up: If you already have good historic product usage data, you can identify these metrics by looking at customers who have churned and retained in the past and seeing which metrics stand out.
Tip: The tendency is to want to track many more metrics. Focus on what you think are the top ones and “park” all other ideas in a separate list. Most metrics highly correlate with each other- look for the ones that distinguish the customer level. Also, carefully consider the time frame for the metric. What time period makes sense for the activity? Should you look at total, peak, median, mean, min?
2. Assign rules for what value for these measures applies to which level
Each Adoption Ladder Level has a rule on how to apply the usage data to assign an account to one of the levels. The definitions should be objective as possible around what “good usage” is, You are not grading on a curve or looking for an even distribution since each product may look very different in profile depending on the product.
Tip: When assigning metric levels, start with L5. What does an exemplar customer look like? You may have very few of these, but you probably know who they are. Then, go to L1: what is the minimum usage of the product that actually delivers real value?
3. Decide on rules for “overall” value across products and modules
If you have multiple products or modules sold to a common buyer, you will typically have one Adoption Ladder for each major product area
Tip: Consider all modules and products the customer has access to and paid for- not just the ones they’ve selected to use. IOW, if they are paying a price for all modules on a platform, their score reflects all modules use- even if zero.
4. Pull metrics, review data and categorizations and adjust
At this point, the product metrics you have selected and the rules you’ve created should be thought of as hypotheses. Once you have pulled the product data and assigned customers to adoption ladder levels, you are not done. The real test is in actually looking at the data and testing how well these hypothetical rules work against actual customers. Iterate on the metrics and ladder definitions until it passes the four tests below. This is critical: proving this out through a series of tests both validates the approach and also builds credibility internally to operate against the adoption levels.
Tip: What you come up with at first will almost always need adjustments- but don’t seek perfection. There will always be outliers and exceptions; there will always be ideas to add or adjust to improve. Aim to get to an “MVP” that is directionally correct and has enough credibility to move forward.
5. Launch as a v1 and then optimize over time as part of future versions of the Adoption Ladders
Just like every product release, the adoption ladders will be a living tool. Over time, you will identify improvements that tighten the correlations and reduce the exceptions. You will have new modules or products that you will want to track.
Most important- like any MVP launch of a product- once it passes the validation tests internally, make the data available and start to operate against it (see below). The metrics underneath the adoption ladders should be reviewed regularly internally as well as with key customers. From this, you will learn what improvements are a priority for future versions.
Operationalizing Adoption Ladders
Once you have established Adoption Ladders and assigned each customer account to them, you now have a way to view how customers may change over time and a framework for how teams can work together to improve adoption. In many ways this becomes the “needle” that needs to be moved.
Product looks for opportunities to improve the product and can look ultimately for improvements in the Ladder mix for that product. These Adoption Ladder metrics can become success metrics upon which new product investments are prioritized and should impact positively.
Customer success organizations aim to be less about “supporting customers” and more about helping customers get the greatest value possible out of the products that they have purchased. With adoption ladders, Customer Success can identify where each account is and develop playbooks for how to move accounts from one level up to another. This may involve encouraging better/increased use of products they already use or supporting initial adoption of products they don’t yet use. It likely includes actions that the company can take to help customers improve the usage of the product as well as tools and tricks that can be shared with the customer so they can support greater use within their organization.
For example, they can work collaboratively with Product for any customer (or internal) training or customer communications that could help move these metrics. Depending on the customer success model, this may involve a combination of scalable activities (eg, webinars, in-app messaging) as well as focusing on specific customers to help them advance on the ladders.
Marketing plays an important role in the messaging and communications that can drive existing customers to higher adoption levels. For example, based on the adoption ladder levels, marketing can more easily identify exemplar customers for writing case studies that can be leveraged with both existing customers and prospects.
Most importantly: all three organizations now have a common language to talk about adoption, to evaluate what “good” looks like, to work together to improve adoption and to measure those improvements over time.
Adoption Ladders in use
The Adoption Ladder framework is adaptable to different businesses and complementary enough to work alongside other frameworks. Launching one at your company is straightforward and offers the ability to optimize over time and a level of flexibility to fit different organizations.
I first developed the Adoption Ladder at a large edtech company whose products were used by one in three K-12 schools in the US. We found that if we could move one single metric (that captured whether a particular methodology was followed in the product) we could move up retention likelihood by 20 percentage points.
My most recent company sells a B2B SaaS product to enterprises and we began using Adoption Ladders as a way to better connect the playbooks of CS with the measurement of adoption. The collaborative process of defining and building the adoption ladders vastly improved the quality of the working relationship between product and customer success teams.
One of my fellow Chief Product Officers implemented Adoption Ladders at his company alongside the Google HEART framework (which also captures subjective scores on “Happiness” and usability measurements around “Task Completion”).
Product-Led Growth strategies fit in well with Adoption Ladders and even know of a fellow product leader is implementing at his B2C company.
At the end of the day, Adoption Ladders don’t replace the product data that good product teams should use to understand how users use their product and how to improve them. They don’t replace the playbooks and tools that CS organizations must create for “customer success.”
They do help key functions in an organization align and operate more effectively together to drive customer success and adoption- arguably the most important business outcomes for any SaaS business.