Case Study Details
Client: Mid-size B2B services company (name withheld under NDA)
Industry: B2B Services
Service: Cross-domain technology diagnosis — AI, Automation, Product, and beyond
Engagement length: 3 weeks diagnostic, 10 weeks execution
The Problem
Growth had slowed, internal teams were stretched thin, and customer complaints were creeping up, but nobody at the company could agree on why, because the symptoms didn't point cleanly in one direction. Their CEO had gotten conflicting advice depending on who she asked.
Her engineering lead thought their product had UX problems driving churn. Her ops manager was convinced the real issue was manual processes eating time that should've gone toward customer experience. A board advisor suggested they were simply behind on AI and needed to "add some automation" without specifying where. Everyone was diagnosing the problem through the lens of their own department, and every proposed fix required committing budget to a specialist team before really knowing if it would address the actual issue.
"We'd gotten three different opinions from three smart people, and all three sounded plausible. That was almost worse than getting no advice at all, we could have spent six figures fixing the wrong thing with total confidence."
That was the actual reason they came to us - not because they knew they needed AI, or automation, or a product redesign, but because they didn't know which one they needed, and every vendor they'd talked to so far was structured to sell them one specific answer before understanding the question.
What We Did First: Diagnose Before Prescribing
We approached this the same way we approach every engagement that starts without a predetermined answer — by mapping the actual business against every plausible bottleneck, rather than starting from whichever service line happened to be asking for the meeting.
Operational walkthrough. We spent time directly with the teams closest to the customer — support, delivery, and account management — tracing exactly where time and effort were going versus where the company assumed it was going. This is often where the real signal lives, because leadership's mental model of "where the friction is" and the frontline team's lived experience of it frequently diverge.
Data and systems review. We looked at what was actually being tracked versus what decisions were being made on gut feel, and audited the tools and systems in place to see where manual work was substituting for something that could reasonably be automated or better designed.
Customer-facing review. Separately, we looked at the product and customer experience itself, independent of any internal assumptions about it, to see whether the complaints pointed to a genuine usability problem or something further upstream.
What we found didn't match any of the three original theories cleanly. The core issue was operational: a manual, error-prone handoff process between sales and delivery that was creating delays and inconsistent customer experiences downstream — which was then surfacing as product complaints, because customers experienced the downstream friction through the product, even though the product itself wasn't the source. The AI suggestion wasn't wrong exactly, but it would have automated a broken process faster rather than fixing it. The product-redesign theory would have polished the part customers actually saw without addressing what was breaking behind the scenes.
What We Built
Because the real bottleneck was a process and systems problem wearing a product-complaint disguise, the actual fix combined two of our disciplines rather than the one everyone had originally assumed:
Automated the broken handoff. We rebuilt the sales-to-delivery handoff as a structured, automated workflow with clear ownership at each stage, removing the manual, inconsistent steps that had been causing delays and dropped details.
Layered in targeted intelligence, not blanket AI. Rather than a broad "add AI" initiative, we implemented a narrow, well-scoped automated check within that workflow to catch missing information before it caused a downstream delay, a small, targeted use of the technology the board advisor had been gesturing at, applied exactly where the diagnosis showed it would matter, not sprinkled across the business speculatively.
Made a small, evidence-based product fix. Once the operational fix was in place, we revisited the original product complaints. A handful were genuine, minor usability issues worth fixing — but a much smaller list than the full redesign the engineering lead had originally proposed, because most of the perceived product friction turned out to be a symptom of the handoff problem, not a separate issue needing its own project.
The Outcome
Within the quarter following rollout:
- The handoff delays that had been driving the original customer complaints were largely eliminated, and complaint volume dropped accordingly
- The team avoided a costly, broad product redesign that would not have addressed the root cause, redirecting that budget toward the operational fix that actually mattered
- Delivery timelines became consistent and predictable for the first time, which the CEO noted was itself becoming a point that their sales team used with prospective customers
- The company came away with a clear framework for diagnosing before spending on their next growth initiative, rather than routing every new problem to whichever internal team argued loudest for their theory
"The value wasn't that they knew AI, or automation, or product design. It's that they didn't assume which one we needed before actually looking. We would have spent our budget on the wrong fix if we'd just gone with the loudest opinion in the room."
— CEO
This case study reflects a real client engagement. The client's name and identifying details have been withheld in accordance with a confidentiality agreement; the CEO's title is used in place of a name for the same reason.