Automating Success: the opportunities and limitations of AI in customer success
Automating Success: the opportunities and limitations of AI in customer success
In our ongoing series on growing your B2B SaaS business from Series A and beyond, we’ve delved into topics such as measuring your dev team’s output and the pivotal role of customer success in driving continuous sales.
Last month, my colleagues Chloe Allan and Rich Bolton provided valuable insights into the critical role of the customer success (CS) team in a B2B SaaS business and shared their guide to setting up an effective one. This month I’m going to build on that foundation and shed light on the role of AI-enabled automation in enhancing customer success.
At Octopus Ventures, we often see companies under investing in CS because of their emphasis on new sales. Often, this results in strong initial traction followed by a growth plateau, as their initial early-adopting customers are left unsatisfied and churn – abandoning the start-up’s solution and looking elsewhere, or going back to doing nothing at all. Newly acquired customers look good on paper, but we know the risks of failing to support them in using your product. That’s why we prefer to invest in companies that can demonstrate a consistent increase in product use, or where a large proportion of the customer base has been through at least one renewal cycle, and it’s why we think customer success is such a vital function to get right.
In a recent article, Customer Success Collective highlighted some of the opportunities for disruptive AI solutions in CS. But their investigation leaves some open questions, such as how far along are we in the automation journey? What has limited AI-adoption in CS? And, perhaps most pressingly, where are we likely to go from here?
In this guide I’m going to unpack the answers to some of these questions – but first, what can AI currently do when it comes to assisting CS staff?
Evaluating AI in CS
As the Customer Success Collective article makes clear, AI can help CS staff to be more efficient. At the same time as allowing them to serve a greater number of client employees, it can assist CS staff in identifying higher value/higher need customers to drive satisfaction, adoption and upsells. In short, AI can:
- Help customer success managers (CSMs) identify upsell opportunities, transforming their role from mitigating churn to proactively driving additional value.
- Assign health scores based on user behaviour to predict customers that are at risk of churning. This allows CS staff to prioritise customers in need of further assistance.
- Triage queries by complexity and urgency, directing customers with generic queries towards serving themselves via chatbots, and freeing up CSMs to answer critical queries and drive upsells.
- Offer recommendations to CSMs on optimal next steps/responses to customers based on a customer’s given circumstances.
Clearly, AI holds the key to unlocking massive efficiencies in CS – so where do we currently stand?
Where are we in the journey?
To get a clear picture, I interviewed Chris Register, Chief Customer Officer of Planhat, a leading CS productivity software from Sweden. Chris explained that the shift towards subscription-based business models over the last 20 years has led to a rapid proliferation of data available for organisations to manage. On top of that, companies now have access to data across multiple formats and sources, including e-mail threads, social media, issue logs, feature requests and survey feedback. All this data can be linked in a continuous way and presents a substantial opportunity for companies that can utilise it. In parallel, as explained in Chloe and Rich’s blog, companies have been pushed to stretch their capital further amidst the challenging macroeconomic context of recent times. Both market drivers have resulted in a heightened focus on the CS function in general.
Despite this, the adoption of dedicated CS tools remains nascent. Dilanka Kalutota, CEO of Octopus Ventures portfolio company, Velaris, told me that most pre-Series B companies that Velaris have sold their solution to have no CS-specific software at all. Dan Steinman, Chief Evangelist, and former Chief Customer Officer, at Gainsight, the largest CS software business in the market, suggested that only around 3000 enterprises globally have adopted a dedicated CS solution. This would imply an adoption rate of well below 1%.
Why is AI adoption in CS still limited?
Dilanka Kalutota, of Velaris, emphasised that leaders in the B2B SaaS sector aren’t aware of the full capabilities of CS solutions and their corresponding value. Before the Series B stage, founders’ priorities lie elsewhere and the need for investment in CS-solutions is often recognised late. But adopting these tools early gives founders the insights they need to develop superior solutions, resulting in an enhanced customer experience and, ultimately, higher net revenue retention.
Companies that have integrated AI into their workflows tend to use it for efficiency enhancing but non-critical tasks, such as content creation for customer discussions and customer health-score rankings. This view was echoed in my discussions with Ben Allen, a Senior CS manager at Flock, an Octopus portfolio company in commercial fleet insurance. Ben said Flock was using AI tools to transcribe meeting notes and summarise customer interactions, but if any new tools were to be considered, they would need to demonstrate an immediate return on investment (ROI) benefit.
During our conversation, Dan Steinman suggested that a few key factors have prevented deeper and more widespread AI adoption in CS. These include the familiar fear of the unknown (i.e., the inability to fully trust AI to respond to customers without human oversight), current budget constraints, lack of clear ROI and a lack of success stories to date. Chris Register, of Planhat, went further, suggesting that the key challenge for early-stage customers was a lack of well-labelled training data for advanced churn-prediction.
Where do we go from here?
As Dan Steinman suggested, we shouldn’t expect a sudden step-change in the adoption of advanced AI in CS, complete with immediate winners and losers. Instead, adoption is likely to be incremental. Customer Success will gradually trend towards becoming more data and process-driven, as well as efficient, measurable, and accountable. AI will be an enabler on this journey, with more CS being completed digitally. Adoption of AI in CS will accelerate as customers are introduced to more AI features through existing customer success platforms. These encounters will give customers the chance to test features and validate ROI ahead of committing to large outlays on expensive additional software.
When we’re analysing new investments at Octopus Ventures, we focus on net dollar retention and how it’s trended across customer cohorts. It’s a sentiment that’s reflected elsewhere, with investors pushing portfolio companies to generate additional value from their existing customer bases. As recognition of the CS organisation’s role in driving additional value grows, we’re likely to see interest in automated solutions follow suit.
The rise of subscription-based models has generated significant data, but the adoption of advanced CS tools has been limited. Companies have struggled to balance a focus on new sales against the need for robust customer management, creating gaps that AI-enabled solutions can fill.
AI has the potential to enhance Customer Success by handling increased customer volumes and prioritising higher-value interactions, but many organizations are still in the early stages of AI adoption. AI tools are layered onto existing processes instead of being used to their full potential.
Hesitancy is attributable to factors like fear of the unknown, budget constraints, unclear ROI and a lack of well-labelled training data for advanced churn prediction. But as CS transforms into a more data-driven, efficient, and measurable domain, industry experts anticipate incremental adoption. As success stories spread, and AI features integrate into existing CS platforms, interest in AI solutions will grow.
With companies recognising the value of an early investment in CS, AI is set to play a crucial role in supporting CS staff as they improve customer satisfaction, retention and upsells. AI adoption might be gradual, but in the long run it has the potential to be a source of competitive differentiation and a driver of meaningful value.
At Octopus Ventures, we, like many other growth-stage investors, encourage proactive investment in CS. A failure to focus on customer success at the right time can harm a business and its investment prospects – but getting it right can yield outsize returns.