Full-stack or mid-stack solutions using generative AI are being launched every week. I alluded to this earlier this week in my post on “Fighter Jets > Copilots.” A question I was asking myself was why are these specific areas coming up as the first ones? What is the market telling us?
A common thing in many of these genAI applications is that they are in areas which have very rational/objective measures of performance, are high volume and are tolerant of some level of lower quality (for a short while, at least). In other words, AI can use private datasets to drive personalized content synthesis/ creation which can then be published, and monitored on objective measures which can in turn drive a better dataset and better personalized content creation. A closed loop.
What are some of these performance-driven high-volume use cases and their corresponding objective measures?
- Performance marketing (click-through rate/ conversion rate/ CAC)
- Search-engine optimization (click-through rate/ leads generated)
- Ecommerce (sales conversion rate, AOV, CAC)
- Feed-driven entertainment (click-through rates, sessions per DAU, share/like rates)
- Gaming (paying user rate, time spent)
- Trading (profit or loss)
- Certain areas of the legal profession (response rate, settlement rate)
- Contact centers (deflection rate)