How the operation itself becomes a system.
Grow is the function nobody owns and everyone needs. It's the difference between an operation that scales and one that just gets bigger. Workforce planning, capacity, hiring standards, AI scope. Across both populations.
v 02 · live
What Grow means.
Grow is the work of turning your operation into a system that scales. Not the work of headcount planning. Not the work of writing job descriptions. The work of designing how capacity, hiring, career paths, team structure, and AI scope evolve together so growth doesn't break what was working.
Most CX teams grow by reacting. A spike in volume; we hire. A senior leaves; we backfill. A new market opens; we cobble together a pod. Workforce planning is a quarterly Excel exercise. Career paths are a wishlist on a Notion page. Hiring standards live in the lead's head.
On the AI side, the parallel problem: scope expands by vendor recommendation, not by operational design. The AI starts taking refunds, then returns, then disputes, with no equivalent of the hiring standard to say "is this within bar." Workforce planning treats AI capacity as fixed when it's actually a decision. Career paths don't account for the fact that the easy work is leaving and the team is being left with the hardest 20%.
That's the gap Grow names. Growth is the most expensive thing a CX operation does, and the least designed. New hires arrive into structure that doesn't match the size of the team. The AI's scope grows by vendor recommendation, not by what the operation can absorb. Senior agents plateau because there's no path. Capacity planning is a fight with finance every quarter, and AI capacity is a fight with no one because nobody is asking the question.
Haven's Grow module starts with naming an owner across both populations. Workforce planning, hiring standards, career paths, capacity model, AI scope: all assigned, all to the same role. Then it builds the artifacts: a workforce model that splits across deflectable and complex, hiring standards that calibrate to the same bar the AI is scoped against, career paths designed for the work humans actually do now.
Growth becomes a designed event, not a reaction. The operation scales as a system, humans and AI in one model. The thing that breaks last is the thing nobody owned first.
The progression. Four levels.
Growth happens to you, on both sides. Volume spikes trigger hiring. Senior departures trigger backfills. AI scope expands by vendor recommendation. Workforce planning is a quarterly Excel exercise. Career paths are aspirational.
- Reactive hiring
- AI scope expanded by vendor, not operator
- Excel-based planning
- Aspirational career paths
Some pieces are owned. AI capacity still isn't. Hiring has a standard. Human capacity has a model. Career paths exist on paper. Nobody owns Grow as a function. It's split across leadership, ops, HR, and (for the AI) the vendor's quarterly business review.
- Human hiring standard exists
- Human capacity model in spreadsheet
- AI capacity treated as fixed, owned by vendor
- No named owner across both
Grow has an owner across both populations. Workforce model splits across deflectable and complex, responds to demand. Hiring standards calibrate to the same bar the AI is scoped against. AI scope changes get reviewed against the same standard, not just the vendor pitch. Career paths map to the seven functions and the work humans now actually do.
- Named owner across humans and AI
- Capacity model splits deflectable / complex
- AI scope expansion reviewed against the standard
- Career paths designed for the harder work
The operation is a self-scaling system, humans and AI in one model. Capacity flexes with demand on both sides. Hiring is continuous, not reactive. AI scope expands or contracts based on calibration data, not vendor recommendation. Career paths produce internal seniors for the work that stays human. The operation grows as a designed organism.
- Flexing capacity across humans and AI
- Continuous hiring pipeline
- AI scope driven by calibration data
- Designed operation, not just a designed team
What Grow builds.
The workforce model
A demand-responsive capacity model that splits across deflectable and complex, sizes the human team for the complex tail, and treats AI scope as an operational decision. Hiring becomes proactive; AI scope becomes intentional.
- 3-, 6-, 12-month capacity forecast across both
- Splits volume into deflectable (AI) and complex (human)
- Adjusts as AI scope changes
- Linked to onboarding ladder
The hiring standard
Calibrated to the bar your operation already holds. Same standard used to scope the AI's expansion, not just to hire humans. Interview scorecards mapped to the competency tiers in Enable. Hiring quality compounds; AI scope decisions become defensible.
- Calibrated to your existing bar
- Scorecards mapped to Enable tiers
- Same standard scopes AI expansion decisions
- Quarterly recalibration
The career ladder
Career paths that map to the seven functions of CX operations, designed for the work humans actually do now (which is, by definition, the hardest 20%). Senior agents have somewhere to go. Promotion criteria are operational, not political.
- Paths mapped to the seven functions
- Designed for the work humans now do
- Operational, not political, criteria
- Linked to Perform & Enable
See it cascade.
Volume trends shift. Grow is where Haven splits the forecast across deflectable and complex, sizes the human team for the complex tail, recommends the AI scope expansion, and adjusts metric baselines so next quarter's numbers compare like for like. One model, both populations, one read. See how the cascade lands in Grow →