The customer picture is scattered across systems, finance and AI tools.
The single most common pain point. Each system holds part of the truth, so the CSM becomes the integration layer.
One CSM managing a portfolio of around two dozen accounts tracks engagement across a CRM, a task board, a proprietary usage tool and email. Another relies on an in-house CRM that lags or shows inaccurate data, which then has to be reported back to developers before it can be trusted. A head of CS cannot see billing information when customers ask, and has to chase finance for answers. An experienced CS builder said they have never met a founder or CEO who did not describe their customer data as "all manual" and spread across many places, calling it "a massive pain point".
The same leader warned that a platform is only "as good as the data that you've got". Integrations were described as "just the biggest pain" of any SaaS rollout, and legacy CRMs as clunky, expensive and in constant need of maintenance.
Companies selling through resellers face a second layer of fragmentation. One leader admitted to having no visibility into how engaged partners are with their own end customers. The team now builds informal support relationships with those end customers directly.
What interviewees asked for was not another dashboard. One CSM described their magic wand as a unified operational view that merges customer metrics, tasks and live activity, then distils them into a value story: actions taken, outcomes reached, and the business case for renewal.
Unify CRM, billing, payments and product usage before trying to model anything on top of them.
The product can be loved and the account can still be lost, because of an invoice.
A CS operations lead said feedback on the product is almost entirely positive. The biggest frustration sits elsewhere: inconsistent auto-pay withdrawal dates and incorrect invoices. They called these high-impact operational problems that quietly erode client trust. Our own research surfaced a parallel pattern: card details expiring on auto-renewing contracts, silently lapsing and knocking customers onto worse rates without anyone noticing.
Billing questions are also among the most repetitive a CS team handles. One head of CS wants billing folded into the customer health view, plus AI-drafted responses to recurring billing and compliance questions that the team could review and send, rather than write from scratch.
B2B SaaS tends to sit at the lower end of the range, but the revenue at stake per account is far higher, and the damage to trust is harder to see.
Billing and payments data belongs inside the customer health view, not in a separate finance silo.
Our free retention health check scores your billing and data visibility. No integrations required.
Check your health score →Appetite for AI is high. So is scepticism about using it as a replacement for judgement or for good data.
A data specialist called it a red flag when people let AI check work with no manual review, recalling an incident where unchecked AI output took a system down for a day or two. When AI is treated as a replacement rather than a tool, debugging and review workload doubles.
A fractional CS leader sees founders assuming AI will "magically" fix scattered data, and said they are "terrified" by how much company data is flowing into AI tools without oversight. Meanwhile, a leader in digital identity pointed to agentic AI as a new risk surface, driving demand for verifiable credentials.
Where leaders did want AI was in removing repetitive work: drafting replies to common billing and compliance questions for a human to review, summarising activity, and surfacing what needs attention.
Use AI to draft, summarise and prioritise inside a governed system, with a human making the call.