8 min read | New research on AI and technology investment on CS capacity
AI Creates Capacity. Operational Maturity Makes It Visible.
Ask a Customer Success leader whether AI is saving their team time, and the answer is almost always yes.
Ask how much time, and where it went, and the answers get much less certain.
That gap showed up in our data too. In ESG’s Customer Success Operations Maturity Assessment, organizations using advanced automation and AI-driven workflows were no better at gauging CSMs’ available capacity than organizations without.¹
So what happens to the capacity AI creates?
Previously I shared that organizations realizing the greatest value from AI had first invested in the operational foundations that make technology effective.
Most AI conversations begin with efficiency. How many hours can we save? How many tasks can we automate? How much administrative work can we remove from a CSM’s day?
Those are reasonable questions. Many organizations are already using automation, AI-assisted communications, digital engagement models, and workflow automation to scale Customer Success in ways that weren’t possible a few years ago.
What many Customer Success leaders fail to recognize is that efficiency gains create expectations.
The moment you tell the executive team that AI can free up hundreds of hours, the next question isn’t about technology. It’s about economics.
What are we going to get for those hours?
If Customer Success has a compelling answer, AI becomes an investment story.
If Customer Success doesn’t, AI quickly becomes a cost reduction story.
And in most organizations, the CFO already has a plan for unused capacity.
Capacity without a reinvestment strategy rarely remains capacity for long.
Capacity is an opportunity. What ultimately matters is whether organizations can identify that capacity, understand where it exists, and intentionally reinvest it into activities that create value for customers and the business.
The data suggest that distinction matters a great deal.
Capacity Created by Improving Process before Adding Technology
In our client work, we regularly see automation reduce the manual effort required to run customer-facing programs.
A digital health software company needed 16 full-time employees to support customer onboarding, and maintaining that headcount wasn’t sustainable as the company grew. After mapping the customer journey and redesigning onboarding, including a shift from one-to-one training to one-to-many sessions, the same work required 8 full-time employees. The team estimated a 60% increase in productivity and could onboard more customers with the same resources.ᵃ
That’s half of an onboarding team’s capacity, freed up through better operational design allowing them to onboard more customers without added headcount and redirect the capacity towards other business outcomes.
Customer Success activities still happened. Automation changed how that work was performed and how much of it a team could carry.
That distinction matters. The aim is capacity that can be redirected toward the customer and business outcomes that matter most.
The Finding That Surprised Me
The first finding challenged one of my assumptions. Going into this analysis, I expected organizations with more advanced AI and automation capabilities to also have a stronger understanding of available capacity.
Even basic automation showed only a modest connection. Organizations automating basic Customer Success motions were nearly twice as likely (1.8x) to understand available capacity.²
In simple terms, implementing AI does not necessarily mean an organization understands how much capacity it is creating.
That observation is consistent with what I’ve seen in practice. Many organizations know AI is helping. Far fewer can answer questions like these:
- How many hours were actually recovered?
- Where was capacity created?
- Which customer segments benefited?
- What happened to that capacity afterward?
That’s where operational maturity enters the conversation. If capacity can’t be measured, it can’t be intentionally managed.
High-Performing Organizations Treat Capacity as an Asset
Organizations that can accurately determine available CSM capacity are three times more likely (3.1x) to demonstrate the ROI of Customer Success investments on Net Revenue Retention (NRR).³
That’s a powerful relationship because it suggests capacity management isn’t really about capacity management.
It’s about credibility.
Organizations that can quantify available capacity appear to be better equipped to answer a question every executive team eventually asks:
What are we getting for this investment?
The pattern shows up elsewhere in the data. Organizations that can accurately determine CSM and CS staff capacity are:
- More than twice as likely (2.3x) to have continuous improvement mechanisms to formally test and refine their engagement models.⁴
- Three times more likely (3.0x) to have a formula for determining the optimal ratio of CS Ops personnel to CSMs.⁵
Taken together, these findings suggest the highest-performing organizations don’t view capacity as an operational metric. They view it as a strategic asset.
As AI creates additional capacity across Customer Success, that distinction becomes increasingly important. Leaders who can quantify and redirect capacity can build a stronger investment case for Customer Success. Leaders who cannot may find that someone else decides how that capacity is used.
Capacity Creates Options. Reinvestment Creates Value.
This is where I believe the story becomes most interesting. The organizations that appear to get the most from capacity reinvest it in better customer outcomes.
Better Customer Engagement
Organizations whose engagement models are easily accessible to CSMs, and consistently adopted across the team, are twice as likely (2.0x) to accurately determine available capacity.⁶
At first glance, that may seem like a finding about documentation. I think it’s really a finding about operational maturity.
AI and automation can accelerate work, but they can’t fix a broken operating model.
If your processes are poorly defined, inconsistent, or disconnected from how customers actually achieve value, you don’t eliminate the problem by adding technology. You simply scale a broken process.
We see this play out regularly in our client work. A digital health software company used cross-functional journey mapping to better understand how customers moved through onboarding and adoption. The exercise revealed that customers were disengaging earlier than the team had assumed. The redesign that followed reduced unnecessary touchpoints, shifted training from one-to-many delivery, and enabled the organization to onboard more customers with the same resources.a
The outcome wasn’t driven by automation alone. It was driven by a clearer understanding of where human engagement created value, where a scaled approach produced the same result, and how work should be redesigned accordingly.
That’s the opportunity. When processes are documented, consistently executed, and aligned to the way customers realize value, technology can amplify what already works.
Put simply: if your processes are poorly defined, you scale a broken process. If your processes are documented, aligned to customer value, and adopted by your team, you scale excellence.
More Effective Risk Management
We documented this firsthand while working with a client. After implementing structured engagement and risk management processes, the organization uncovered that 42% of annual contract value was at risk, much of it among accounts that had received little or no proactive engagement.b
Findings like this are why I believe one of the most important questions in Customer Success isn’t “How much capacity can AI create?” It’s “Where should we reinvest the capacity we create?”
AI and analytics can surface warnings. Capacity gives teams the ability to investigate, intervene, and change the outcome.
The data reinforces that point. Organizations with documented risk management processes are nearly three times more likely (2.7x) to demonstrate the ROI of Customer Success costs on NRR.⁷
In other words, the value isn’t in identifying risk. The value comes from having the operational maturity to do something about it.
Cross-Functional Enterprise Value
One finding I found particularly interesting involves organizations that document the value Customer Success provides to other business functions. Those organizations are:
- More than three times as likely (3.6x) to use digital and AI metrics to inform engagement strategy.⁸
- Three times more likely (3.0x) to have a formula for determining the optimal CS Ops-to-CSM ratio.⁹
- Twice as likely (2.0x) to demonstrate the ROI of CS costs on NRR.¹⁰
My read is that the most mature organizations reinvest capacity beyond Customer Success itself. The survey doesn’t measure that directly, but in my experience those organizations contribute customer intelligence, expansion opportunities, and market insight that inform decisions across the business.
What Doesn’t Matter As Much As We Might Think
One advantage of quantitative analysis is that it challenges assumptions as well as confirming them. Several relationships many leaders might expect to see were weak or missing.
As noted earlier, neither advanced nor basic automation showed a strong connection to understanding capacity.¹ ² Both were far weaker than the relationships tied to operational practices.
Perhaps the most surprising result: seeing your capacity clearly doesn’t give you more of it. Organizations that could accurately determine CSM capacity were just as likely as everyone else to say their CS Ops team is understaffed.¹¹
Anyone who has built a capacity model knows the feeling. The model shows you exactly where the time goes, and it can still tell you the team is stretched. What visibility gives you is the ability to make the case for resources, or to decide with confidence what to stop doing.
The organizations creating value appear to share a set of operational practices that help them understand and manage capacity intentionally, regardless of headcount or level of automation.
Final Thoughts
Many organizations begin their AI journey with a simple question: How much time can we save?
The organizations creating the most value eventually ask a different set of questions.
- Where is capacity being created?
- How much capacity exists?
- Where should it be reinvested?
- How do we know that investment improved customer outcomes?
AI may be one of the most important capacity creators Customer Success has ever seen. Operational maturity is what turns that capacity into enterprise value.
The organizations realizing the greatest return from AI will be the ones making the most deliberate decisions about how that time is reinvested.
Those decisions start with knowing where your capacity sits today. If you’re ready to answer the questions above for your own team, our Tiering and Capacity Planning Workbook is a practical place to begin.
Download the Tiering and Capacity Planning Workbook
Notes and sources
Source: ESG Customer Success Operations Maturity Assessment, responses from over 200 organizations. Each comparison includes only organizations that answered both questions. “More likely” shows how strongly the presence of one capability was associated with the presence of another. Findings 2-10 are statistically significant (Fisher’s exact test, p < .05) and remain significant after correcting for multiple comparisons (Benjamini-Hochberg, q < .05). Findings 1 and 11 are not statistically significant. Findings describe association, not causation.
- Use of advanced automation and AI-driven workflows to scale CS motions, and the ability to accurately determine CSM and CS staff capacity: 51% (26 of 51) vs. 39% (41 of 105). Not a statistically significant difference.
- Automation of basic CS motions with email campaigns and/or AI-assisted communications, and the ability to accurately determine CSM and CS staff capacity: 55% (42 of 76) vs. 31% (25 of 80).
- Ability to accurately determine CSM and CS staff capacity, and the ability to demonstrate the ROI of CS costs on net revenue retention: 49% (33 of 67) vs. 16% (14 of 89).
- Ability to accurately determine CSM and CS staff capacity, and a formal process for testing and refining engagement model effectiveness: 31% (21 of 67) vs. 13% (12 of 89).
- Ability to accurately determine CSM and CS staff capacity, and a formula for the optimal ratio of CS Ops personnel to CSMs: 30% (20 of 67) vs. 10% (9 of 89).
- Engagement models easily accessible to and adopted by CSMs, and the ability to accurately determine CSM and CS staff capacity: 56% (47 of 84) vs. 28% (20 of 72).
- Documented risk management process to identify and escalate at-risk customers, and the ability to demonstrate the ROI of CS costs on net revenue retention: 44% (34 of 77) vs. 16% (13 of 79).
- Documented value CS provides to other business functions, and digital CS and AI-enabled operational metrics informing engagement strategy: 56% (41 of 73) vs. 16% (13 of 83).
- Documented value CS provides to other business functions, and a formula for the optimal ratio of CS Ops personnel to CSMs: 29% (21 of 73) vs. 10% (8 of 83).
- Documented value CS provides to other business functions, and the ability to demonstrate the ROI of CS costs on net revenue retention: 41% (30 of 73) vs. 20% (17 of 83).
- Ability to accurately determine CSM and CS staff capacity, and CS Ops headcount adequate to support the entire CS organization and customer base: 31% (21 of 67) vs. 30% (27 of 89). Not a statistically significant difference.
Client examples come from ESG’s published case studies. They are reported by ESG and its clients and are separate from the assessment analysis above.
- Digital health software company: Transforming Onboarding for Enhanced Efficiency, Improved Adoption, and Increased Retention. Results are from a four-month pilot.
- Identity security company: Improving Retention and Enterprise Value Through a Scalable Customer Success System. The at-risk ACV figure applies to the pilot segment only.

