The Proprietary Technology Speed Advantage
Own critical technology so customer improvements move on your timetable
- Difficulty
- Expert
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 97%
Cohan argues that relying solely on third-party technology constrains both what a company can develop and when it can deliver improvements. His alternative is to own the critical technology layer, then use internal development to automate workflows and customise solutions around customer needs. AI enters as another production capability on the back end, accelerating work that the company already understands rather than serving as an isolated experiment. The method starts with a customer delay or limitation, identifies the technology dependency causing it, and brings the strategically important capability under direct control. The company then automates the workflow, adds AI where appropriate, and measures faster delivery of the finished product. The intended advantage is not ownership for its own sake, but control over development speed, customisation, and customer experience.
Origin
Cohan connects Unisource's current proprietary automation and AI work to his long history of building credit-reporting technology internally.
Core principles
- 01Critical development speed should not depend entirely on a vendor
- 02Owned technology enables customised customer solutions
- 03Automation should shorten delivery time
- 04AI extends an existing automation strategy rather than replacing it
How to run it
- 1
Locate the constraint
Identify the customer-facing improvement that cannot move quickly enough because a third party controls the development schedule. Describe the delay in operational terms.
Pro tip Prioritise recurring bottlenecks that directly affect delivery time or customer experience.
Watch out Do not build internally merely because ownership feels prestigious.
- 2
Choose what to own
Bring the critical capability under direct control while leaving commodity functions to external tools. Match the scope to developments and time frames that create real differentiation.
Pro tip Own the layer where customisation and speed matter most.
Watch out Trying to replace every third-party system creates unnecessary maintenance.
- 3
Automate the workflow
Build the proprietary process around the complete customer outcome rather than a disconnected technical feature. Remove avoidable manual delay from production.
Pro tip Use internal developers who understand the industry's operating needs.
Watch out Automation that accelerates the wrong workflow does not improve service.
- 4
Add AI selectively
Apply AI on the back end where it can help produce a finished product faster. Keep verification and industry requirements attached to the workflow.
Pro tip Extend proven automation before inventing an unrelated AI use case.
Watch out A faster output is not useful if its quality cannot be trusted.
- 5
Measure customer speed
Compare the new delivery time and experience with the constrained process. Continue development where the owned capability produces a meaningful customer advantage.
Pro tip Track the finished product's turnaround time, not only internal model speed.
Watch out Technical progress without a better customer outcome is not the intended advantage.
In the wild
Unisource combines proprietary development, automation, AI work, underwriter relationships, and external developer relationships to shorten title-report production. Cohan says the company can already deliver full instant title reports in some states within three, four, or five hours.
→ A traditionally slower service is delivered to customers within hours in supported states.
Cohan's earlier company built its own credit-report logic and a centralised platform used by about 120 credit agencies. Internal development and integrations helped customers obtain reports rapidly through loan-origination software.
→ Owned technology supported faster report delivery and enabled other agencies without comparable systems.
Common mistakes
Building everything internally
The advantage comes from controlling critical differentiated capabilities, not rejecting every useful third party.
Starting with AI rather than the workflow
AI should accelerate an understood production process and customer outcome rather than exist as an isolated feature.
Measuring technology instead of service
Development activity matters only when it improves delivery speed, product quality, or customer experience.
Is it for you?
Best for
It is best for established technology-enabled service firms with recurring workflows and the resources to maintain proprietary systems.
Not ideal for
It is not ideal for early businesses whose standard needs are already met cheaply by reliable third-party software.
From the transcript
“you stifen yourself if you're solely relying on third party Technologies”
“if you're relying on solely a third party to get that work done for you and you're working on their time schedule it's going to…”
“we're also implementing AI where on the back end it's going to be able to help us produce at a faster speed to the customer”
From the episode
Revolutionizing Mortgage ft. Pioneer Michael Cohan
Pioneer Michael Cohan