Mature artificial intelligence (AI) customer programs view customer retention as a product and report customer retention (NRR) between 118%-125%.
Manual operations retain customers at a stagnating 98% to 105%. As of this date, 78% of software organizations utilize algorithmic tools throughout their post-sales workflows to reduce the amount of time to first value by 42% and to identify accounts at risk of churn 34% faster.
The following guide will provide an overview of the actual formulas that make up the segmented benchmarks and the automated structures required to move software retention from a reactive support function to a mathematics-based revenue engine.
The Baseline for Customer Retention Revenues
The industry is heavily dependent on outdated and vague benchmarking that has become an industry standard. Most companies provide numerical retention metrics with a goal of 100% net revenue retention or 90% gross revenue retention for all of their customers. Due to the broad and generalized nature of this goal, the majority of companies have found this benchmark to be useless and non-applicable.
The retention metrics for different market segments and average contract values will not scale the same across all verticals. A single median number will mask the true structures for the way that different customer cohorts behave. To build an accurate customer retention operating model, management must segment their customer retention goals based on the specific type of customers that they acquire.
Setting Segmented Targets
The average NRR for a median private B2B software company will be approximately 106%. If you dig into the data, you will see very strong differences between market segments. Enterprise-level businesses have a typical NRR of 118%, and mid-market companies are usually around 108%. Small and medium business-focused companies (SMBs) are typically below 97% NRR.
If the company has an NRR below 100%, they are shrinking and losing more revenue to churn than they are gaining through expansion. Best-in-class companies create their customer retention revenues beyond 130% NRRs. Achieving high growth numbers for a company means that you will need to implement a rigid segmented strategy. A company's enterprise engagement model cannot be applied to an SMB client base as the unit economics of each segment will collapse.
Understanding Gross Margins
Gross revenue retention (GRR) quantifies how much of the original revenue you retain without considering upgrades or expansions. It provides the most unrestricted view of the product's value since this will include all current customers’ expectations. The median gross retention for the industry averages 90%.
Generally, a good GRR falls between 90%-95% with excellent programs exceeding 95%. For management of SMB clients, due to the increased rate of business failure or cancellation of their software subscriptions, an 85% GRR is generally deemed acceptable. The industry's average annual revenue churn rate is about 3.8%. B2B sectors average slightly above 4.9%. For early-stage companies generating less than $300K annually in recurring revenue, the monthly churn rate can be as high as 6.5%. That number often stabilizes to around 3.1% once the company has crossed the $8 million mark in recurring revenue.
Setting Up the Onboarding Process
The onboarding process is a critical point in the customer journey and a core part of effective customer success strategies for saas. Many organizations view onboarding only as a technical setup and do not leverage this phase effectively. It needs to be designed to provide the customer with their primary value as quickly as possible.
Improving Time to First Value
Time to first value (TTV) is the most important metric for early retention of customers. For mid-market and enterprise customers, you should target to achieve TTV within 30 days. Otherwise, the customer could lose both momentum and internal support for the product.
To create a standard operating procedure (SOP) for onboarding, it is necessary to define the five stages of the onboarding journey:
Defining the goal
Connecting the various systems
Establishing the special environment required for onboarding
Providing support for end users once they start the onboarding process
Celebrating the success of the first user
Algorithmic onboarding solutions have dramatically reduced the time needed for an organization to reach its onboarding time value (TTV) with an average decrease of 42% and an increase in activation rates (AR) of 31%.
Using Progressive Disclosure for Adoption
Once the initial onboarding method has been completed, the relationship between the customer and the organization does not end. One of the most significant failures is that after the first 30 days, customers tend to stop using the solution. To continue to support, educate, and encourage regular use after the first 30 days, companies will need to plan how they will implement a systematic approach to ongoing adoption.
A key factor in this process is progressive disclosure. This means you do not provide all the features of the system to your users in one step. Features are mapped to expected outcomes and introduced to users based on their ability to efficiently utilize the previous tier of functionality. Using this approach creates a positive experience for users learning the fundamentals of the solution and provides opportunities for growth.
Creating Accurate Health Scores
Most health scores are worthless because they are not based on objective data to evaluate the success of a solution. Instead, they are based on subjective color coding based on how the account manager feels. An actual health score must be a combination of mathematical constructs. These include how frequently a customer utilizes the product or service, the sentiment of support tickets associated with the product, and measurable customer engagement metrics to provide a numeric probability of renewal.
A health scoring matrix should be created from available data by using specific weights to make health scoring formulas. You cannot verify health using only the total login count. You need to know what specific activities were performed by that user that directly correlate to the value the organization received for the customer's business.
Core feature adoption (40% weight): Monitoring the levels and amounts of use of the three most important platform features. Should usage drop below a predetermined level, this indicates an immediate possible risk.
Support ticket sentiment (30% weight): Evaluation of the words and volume of support requests. A large number of requests is not necessarily a problem; angry words and recurring bugs indicate a greater chance of customer loss.
Executive engagement (30% weight): Date of last contact with the executive buyer. If the person signing the check hasn’t attended the last review meeting for the past 6 months, the account is in jeopardy despite the volume of users using the product daily.
Setting Up Alerts and Service Rules
A scoring matrix is only beneficial if it triggers an immediate reaction. If the health status for an account drops significantly below the baseline, the system should notify the person responsible for that account.
There are strict rules related to SLAs. If an account has shown a 30% decrease in core feature adoption over 14 days, the manager must initiate contact within 24 hours. Algorithmic scoring tools have an advantage here. Predictive health scoring will identify at-risk accounts 34% faster and generate a 28% higher success rate of retaining that account versus manually monitoring.
Creating Action Plans
Action templates outline what each employee should do in the event a defined condition is met. You cannot expect staff to improvise solutions when an account is at risk or there is the potential for an expansion opportunity. The process must be developed, validated, and standardized.
Steps for Low-Engagement Recovery
When users stop using your product, the system must be able to implement a series of recovery options based on your data. When vague commands are issued, such as “reach out to the customer”, they will not be effective.
The framework must include triggers, channels, and timing indicators. For example, if a customer's active usage decreases by 30% within two weeks, the software sends a targeted email about a highly relevant feature. Two days after the email, an in-app prompt is delivered. If the customer does not respond within five days, a task is assigned for the account manager to call the customer's primary contact. Each of these sequence effectiveness must be tracked using clear recovery metrics to optimize your customer success strategies for saas.
Understanding the Renewal Process
You should not surprise customers with their renewals. Customers should have knowledge of how the renewal process will work months before they are contacted by the software company. One of the most significant failure areas for many software companies is that the post-sales team and sales department do not have a mutual understanding of who is responsible for the commercial transaction at the time of the renewal.
You must have clear rules of engagement in place. In some companies, the post-sales manager manages all renewals, while in others, they transfer the account back to the sales representative. A responsibility matrix is required to specifically outline who is accountable for the commercial negotiation and who will handle the technical success planning. Properly structured compensation plans must ensure that bonuses are tied to net revenue retention (NRR) and gross revenue retention (GRR) outcomes.
Artificial Intelligence in Post-Sale Processes
Algorithmic approaches to managing post-sale processes are no longer just a future possibility. This technology is broadly used in software companies, and as a result, a high percentage (78%) of software organizations incorporate algorithmic technology into their workflows. Automated scoring and identification through artificial intelligence has become one of the most powerful automated solutions to help SaaS organizations make faster and smarter decisions.
The combination of using customer success platforms (CSPs) such as Gainsight, ChurnZero, and Totango means a company now has the ability to analyze large amounts of data to identify patterns that human agents are likely to miss. These platforms have integrated machine learning capabilities into their various functions and processes.
The impact of this is great. Companies using algorithmically driven CSPs have reported a 15-25% reduction in gross churn within one year. These types of platforms analyze product telemetry, billing history, and communication history to create predictive models based on the information they receive. A human is typically not able to know a customer is unhappy until after the customer has made the decision to leave the business. Algorithms will be able to identify subtle changes in a customer's behavior before they make the decision to leave.
Customers are now more demanding of personalized engagement than ever, and manually scaling personalization can be difficult if not impossible. Automated systems are now able to automatically provide personalized guidance for customers. For example, if a user attempts a specific workflow and experiences a delay, the platform will automatically send them the documentation or in-app instructions necessary to resolve the dilemma.
The top 500 B2B software companies that have implemented these platforms report significant improvements:
A 15 point increase in net revenue retention (NRR).
Improved gross retention (GR) from an average of 90% to 95%.
The amount of time needed to prepare for quarterly business reviews has decreased by an average of 60% due to automated data aggregation.
Revenue generated through the expansion of products has increased by 30% due to the automated identification of signals.
Using this technology will require management to create a change management plan since employees often are resistant to the new automated tools. Management must provide employees with a detailed checklist of how to implement the automated tools and conduct training to move from manual data entry to strategic execution.
Planning Team Capacity
The optimal staff capacity must be calculated to know the maximum number of sequence executions that can occur on a certain number of accounts. This calculation is based on the mathematics of how many account loads a single person can effectively manage before service quality begins to deteriorate.
Managing Account Coverage Ratios
Ratios of coverage are determined by the complexity and the market segment of the product. Having an individual with 200 accounts assigned to them guarantees downward spirals.
An enterprise model with high-touch design usually is 1 to 10 and can range up to 1 to 30. These individuals deal with large contracts, as the product requires more extensive and specific strategic planning. The average yearly revenue per enterprise representative averages between $2 million and $5 million of annual recurring revenue.
Mid-market style operations typically take a ratio of 1 to 30 up to 1 to 75. Moderate levels of interaction are needed with these accounts as opposed to a high degree of unique tailoring. Low-touch models, typical of a product-led growth (PLG) model, use ratios of 1 to 100 to even as high as 1 to 500+. This level of scale is made possible largely through extensive use of in-app messaging, automated action sequences, and self-serve documentation.
Aligning Revenue Across Departments
Post-sale teams do not solely have the responsibility of retaining customers. Churn will also result when the product team allows for the development of features not based on customer needs, and the sales team sells to the wrong target profile.
Empirical evidence indicates that when sales, product, and post-sale departments align their metrics and communications perfectly, net retained revenue (NRR) increases as much as 22%. The combined cost of all three departments should generally not exceed 10% of total annual recurring revenue. Remaining below this threshold is a key component of viable customer success strategies for saas, helping ensure the operation remains a profitable revenue-driving engine.
The Science of Post-Sale Revenue
Gone are the days of managing software customer retention through subjective customer relationships and generic advice. The current marketplace relies on a mathematical and statistical methodology-driven model. The data shows companies that employ algorithm scoring, enforce strict time to first value (under 30 days), and have comprehensive documented intervention methods have better financial performance.
Companies have changed their operational bases from just preventing cancellation to systematically engineering a reliable and increasing revenue stream. No longer is success in the enterprise sector an art, but rather a science based on rigorous data analysis.
