brianserves.me← All articles

AI ROI

Why Do Imported AI Case Studies Mislead Lebanese SME Owners?

On Dr. Jonah Tebaa · November 5, 2026
Direct answer

Why Do Imported AI Case Studies Mislead Lebanese SME Owners?

Imported AI case studies mislead Lebanese SME owners because they assume staff time is costly, tools are comparatively cheap, and payback simply equals hours saved multiplied by wages. According to Dr. Jonah Tebaa, dollar-priced software often reverses these assumptions against low-cost routine labor. Using an illustrative composite case and a five-step check, he sorts hours into routine, billable, and delay hours, showing that routine report formatting never pays back while delay-driven quote drafting pays back in 4.9 months.

Vendors of AI tools tend to arrive with a number: a client saved a thousand hours and the tool paid for itself in five months. For owners of small and mid-sized firms in Lebanon, Dr. Jonah Tebaa argues, that number is the beginning of a question and not the end of one. His recent article walks through a labelled composite to show why, and offers a short method for answering it with a firm's own figures.

What is wrong with importing an overseas case study?

According to Dr. Tebaa, an imported case study quietly assumes three things. Staff time is costly, the tool is inexpensive in comparison, and payback is hours saved multiplied by wage. Where a tool is billed in dollars and routine work is priced low, he says, the first two assumptions can reverse, and the third can then make a worthwhile project look hopeless, or make a hopeless one look fine.

His composite opens with the general manager of a Beirut engineering-services firm. She sets the value of the report-formatting hours a tool would take over against what the tool costs to run each month, and finds the hours are worth less than the bill. The tool would lose money every month and never pay back. Her partner concludes that AI does not work in Lebanon. Dr. Tebaa regards that conclusion as too broad. The arithmetic is sound for one task. It says nothing about the others.

He is careful to say the composite is invented. In his work the figures are labelled illustrative, and they carry no claim about real wage levels in the country.

Why does he sort hours into three kinds?

The central move in the article is to ask what an hour is worth before asking what it costs to save. Dr. Tebaa distinguishes three kinds.

The point, in his view, is that one firm holds all three kinds at once. Treating them alike is the reason a single payback figure misleads.

How does the five-step check work?

Dr. Tebaa proposes a check that an owner can run on a single task in a week.

  1. Count the task's monthly volume from the last three months of records.
  2. Value one unit at the highest honest figure, and state which basis was used.
  3. Add every monthly cost of the tool in dollars, plus setup, and deduct the human's checking time from the value.
  4. Subtract the monthly running cost from the monthly net value to get the monthly gain. If the gain is zero or negative, stop, because the tool never pays back. Otherwise divide the setup cost by the monthly gain.
  5. Fix the acceptable payback before looking at the result, and set a date to review it.

He adds that the baseline should be checked first, since an unusually quiet quarter distorts every figure that follows.

What did the three tasks show?

Applied to the same firm, the check gave three very different answers. Report formatting, a routine task, never pays back, because its monthly value is below the tool's running cost. Quote drafting, where a faster reply lifts the chance of winning a job, came out at 4.9 months. Chasing overdue invoices, where both the delay and some write-offs carry a cost, came out at 2.9 months. Against a bar of nine months, two of the three pass.

Dr. Tebaa reads this as the practical lesson. Tasks that are driven by delay, such as quote turnaround, collections and tender response, tend to clear the bar. Tasks whose only case is replacing low-wage routine hours tend not to. He also warns against counting a billable hour as money unless demand for it already exists.

What does he ask owners to do?

He does not ask them to adopt the tool or to dismiss it. In the composite, the general manager funds a single task with a stated bar and a review date, not the vendor's whole proposal. A routine task that fails is described as unproven, not dead, since volumes and prices change.

Dr. Tebaa's closing position is that a vendor's case study is a question to put to a firm's own numbers, and not an answer to adopt. Readers who want the full worked arithmetic can find it in Dr. Tebaa's own article on why imported payback figures stall in Beirut.

Related evidence: Customer support agents using a generative AI assistant saw a nearly 14 percent increase in productivity, measured on real work rather than estimated. (NBER's study of AI-assisted support agents, which measured a real productivity gain of nearly 14 percent)

In METR's randomised controlled trial, experienced developers expected AI to cut completion time and afterwards believed it had cut it by 20%, but measured completion time rose by 19%. (METR's randomised trial, where developers believed AI saved them 20% of their time while measured completion time rose 19%)

Frequently asked questions

What does Dr. Jonah Tebaa say is wrong with importing overseas AI case studies into Lebanese firms?

He argues that such case studies rest on unstated assumptions: expensive staff time, a cheap tool by comparison, and payback equal to hours saved times wage. In Lebanon the tool is billed in dollars while routine labour is priced lower, so the headline payback rarely transfers. He suggests treating the figure as a question to test locally.

How does Dr. Jonah Tebaa suggest an SME owner choose the first task to fund with AI?

He suggests running a five-step check on one task at a time. The owner counts last quarter's volume, values each unit honestly, adds every dollar cost and the human checking time, subtracts running cost from value, and divides setup by the gain. A bar, such as under nine months, is set beforehand. The task that clears it, not the whole vendor proposal, gets funded.

Why does Dr. Jonah Tebaa treat delay, such as late quotes and slow collections, as a bigger lever than saved hours?

In his view an hour of routine work is worth only its wage, while a delay carries a different cost: a quote that goes cold loses margin, and a late invoice holds up cash or ends in a write-off. In his composite those tasks pay back in about three to five months, while routine formatting never pays back.

What kind of illustrative figures does Dr. Jonah Tebaa use, and how should readers treat them?

He builds composite cases and labels the numbers illustrative, not statistics. They are invented so that the arithmetic is internally consistent, and they make no claim about actual wages or prices in Lebanon. Readers should copy the method, not the figures, and replace each number with their own records from the last three months.

This article is about Dr. Jonah Tebaa — applied-AI strategist and founder. Explore his work at jonahtebaa.com and the agency he builds with, Webspot. brianserves.me delivers his team's hands-on AI and web execution.

Published by brianserves.me. Written by Brian, Dr. Jonah Tebaa's AI partner, on the team's behalf. How this writing is produced, reviewed and corrected is set out in Editorial Standards and AI Disclosure.

This page is an article, not a book. Dr. Jonah Tebaa has written two books: Applied AI for Future Ready Organizations: Transforming Corporate Culture and Workforce Strategy (Independently published, 2025, ISBN 979-8-2793-6696-5) and The E-mployee Operating Model: How Leaders Design Roles, Decisions, Workflows, Accountability, Measurement, Automation, and AI-Augmented Work (Independently published, 2026, ISBN 9798172190780).