A Critical Perspective on AI ROI
- A recent issue (#129) of Goldman Sachs’s Top of Mind Report paints a gloomy picture of AI investments, focusing on large tech companies’ struggles to show immediate returns. This view is myopic and fed inside an executive echo chamber. It overlooks the broader economic impact and the tens of thousands of documented successes of small and large organizations and individuals leveraging AI.
- However, we continue to see a trend of enterprises struggling with wide AI adoption of mainstream tools like ChatGPT Enterprise or Microsoft Copilot, derived from 8 of 10 employees lacking skills to get the full value from those tools.
Industry Tremors and AI ROI Realities
- The Grassroots Advantage: Our data consistently shows 10x to 40x ROI in companies using generative AI, primarily in SMBs and among individuals at large enterprises. The high ROI comes from automating labor-intensive activities such as support, content creation, personalized outreach, and content/data summarization and analysis.
- Large enterprises face an “efficiency trap”: AI-driven time savings benefit employees without boosting organizational productivity. Employees maintain the same salary while working less, creating an “efficiency leak.” Large enterprises struggle to redirect these efficiencies, unlike SMBs and independent consultants who reinvest AI-saved time into more productive tasks. How can enterprises ensure that time saved through AI translates to bottom-line improvements?
The Gen AI Pricing Paradox, Race to the Bottom vs the Profitable AI Creators (ISVs) Economy:
- GPT-4o-mini: Our lab tests show that OpenAI’s latest release, GPT-4o-mini, procured from Azure, is 37% faster and 40 times cheaper than GPT-4o with 78% more consistent response times. This continues a price reduction trend that many attribute to pressure from Open-Source models such as Llama 2; however, the root cause is a failed attempt to adapt software and SaaS pricing models to something as disruptive as Generative AI.
- We reiterate our forecast of significant shifts in pricing models. Ten percent of AI creators are embracing pricing models based on the economic benefit of automating equivalent labor.
- Low-code apps, wrappers, and single-purpose AI tools continue to dominate: Single-purpose AI tools continue to dominate the market in adoption, satisfaction, and profitability. AI creators (ISVs) pay an average of $0.0045 per AI request but sell outputs at a staggering 144x margin. Even at that high margin, their clients report savings of 50% or more and are delighted.
Trust and Transparency: The New AI Imperatives
- Trust at an all-time low: Our July tracker shows an all-time high of 79% of individuals concerned about AI privacy and security.
- We’ve launched the AI Transparency Notice initiative.
- 90% support the initiative, wishing all AI vendors would adopt it.
- Transparency = Sales:Â 67% would trust and use (i.e., buy) AI tools more with such notices.
Resources
- ​Workshops – Inform strategies and get teams aligned
- ​Copilot Ada – AI Learning Coach for your Organization
- ​AI Compass – Free Personalized Plan to Adopt AI
- ​Skills Compass – Individual Benchmark & Report
Luis J. Salazar
Founder | AI4SP
Sources:
Our insights are based on data from over 125,000 individuals and organizations who used our AI-powered tools, participated in our panels and research sessions, or attended our workshops and keynotes.
