Three-Bucket Technology Investment Test
Fund technology that scales work, strengthens people, or improves convenience
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 98%
Every proposed technology investment must produce at least one of three outcomes. It should automate manual work to generate economies of scale, make employees more capable when serving customers, or make the customer's interaction more convenient. Leaders first identify the workflow and classify the proposal, then define a measurable baseline such as handling time, employee effectiveness, error rate, capacity, or customer effort. A limited pilot tests whether the technology changes that measure before broad deployment. The framework keeps technology subordinate to operating objectives while still allowing overlap between buckets. Its intended output is sustained operating leverage: the organization serves more customers and manages more activity without requiring a proportional increase in personnel.
Origin
Extracted from The Shawn Ryan Show
Core principles
- 01Automate repeatable manual work
- 02Use systems to augment employees
- 03Remove friction for customers
- 04Connect technology spending to an operating outcome
- 05Scale service without scaling headcount proportionally
How to run it
- 1
Identify the workflow
Specify the manual, employee-assisted, or customer-facing process the investment will change.
Pro tip Observe the current workflow before designing its replacement.
Watch out Do not begin with a fashionable technology searching for a problem.
- 2
Choose the primary bucket
Classify the proposal as automation, employee augmentation, or customer convenience.
Pro tip Record secondary benefits, but require one primary outcome.
Watch out A project that fits no bucket may lack a clear operating purpose.
- 3
Establish a baseline
Measure the current cost, capacity, error rate, service quality, or customer effort.
Pro tip Use a metric directly affected by the workflow.
Watch out Without a baseline, efficiency claims cannot be verified.
- 4
Design the new interaction
Define how work, decisions, and handoffs will change for employees and customers.
Pro tip Include frontline employees in workflow design.
Watch out Automating a broken process can magnify its defects.
- 5
Run a bounded pilot
Test the investment with a limited user group and predetermined success criteria.
Pro tip Track both efficiency and experience metrics.
Watch out Short pilots may conceal maintenance or adoption costs.
- 6
Scale proven value
Expand the system when evidence shows a durable improvement over the baseline.
Pro tip Reinvest part of the savings in further workflow improvement.
Watch out Do not assume pilot performance will automatically persist at scale.
In the wild
A financial firm gives representatives a unified system showing client history, portfolio context, and pending requests during calls. It pilots the system with one team and compares resolution time, repeat contacts, and satisfaction with the old workflow.
→ Representatives resolve more requests while delivering a more consistent experience.
An operations unit replaces repetitive spreadsheet reconciliation with rules-based matching and routes only exceptions to specialists. Leaders measure processing capacity, errors, and staff time before expanding the system.
→ Routine volume scales while employees concentrate on difficult exceptions.
Common mistakes
Buying technology without a workflow
A tool cannot create value when leaders have not identified the process it will improve. Start with the operating problem.
Measuring adoption instead of outcomes
Usage alone does not establish efficiency or service improvement. Compare operational results with a preimplementation baseline.
Removing necessary human judgment
Some interactions require expertise and empathy. Use technology to strengthen people when full automation would damage the experience.
Is it for you?
Best for
It is best for leaders prioritizing digital investments in service-oriented, operationally complex organizations.
Not ideal for
It is not ideal for foundational research whose value cannot yet be tied to automation, employee capability, or customer convenience.
From the transcript
“First, um, create automation where automation is possible, so you can generate economies of scale for your investors.”
“Second, can you use technology to make your people better”
“And then the third is how do you make it more convenient for your investors.”
From the episode
Interview With William McNabb: Masters in Business (Audio)