The challenge
Dandelion's product pairs physical hardware with monitoring software, making for an inherently complex sale before workflow problems even entered the picture. A previous Salesforce implementation had been left unfinished, leaving the internal team distrustful of the platform, and proposals, scheduling, and follow-ups were all handled manually — a source of delay and error.
As Dandelion expanded to serve developers and large-scale housing projects, the system stalled outright: a single B2B proposal could take up to two months to complete. Frequent turnover among the internal Salesforce focal points compounded the problem, leaving the system without a consistent administrator.
The work
Laminar stepped in as a strategic Admin/Support Squad, centralizing sales and support on Sales and Service Cloud. The core decision was Salesforce CPQ — automating Dandelion's complex, hardware-plus-software pricing logic to remove manual error and cut proposal turnaround time, at the cost of the upfront work needed to model that pricing logic into the system.
Field Service was layered in to manage physical components and on-site installations, and Account Engagement to nurture both B2C leads and B2B construction partners, backed by flow automation for stage-change alerts, contract generation, and integrated email follow-up sequences. The new objects and automations went in without disrupting Dandelion's current-state operations — a constraint that mattered as much to restoring trust in the platform as the features themselves.
The impact
B2B proposal turnaround time dropped from two months to one week. In a single week, the team generated seven complex proposals — a 56x acceleration in the pricing process. Conversion rates rose 30%, driven by automated follow-ups and faster proposal delivery, and total revenue rose 20%.
Just as important to the account: the new objects and complex automations went in without disrupting existing operations, restoring the internal team's confidence in Salesforce after the earlier, unfinished implementation. All figures are operator-reported.