The Revenue Engineering Library
Open questions
What the method cannot answer yet. For each: why it matters and what evidence would settle it. If you have that evidence, write to me.
Reviewed quarterly · last review 29 September 2026
Open
Does the cannibalization level at which a new branch stops paying differ between Riyadh and smaller cities?
- Why it matters
- The opening gate counts only the sales a new branch adds, not the sales it takes from its siblings.
- What would settle it
- Before/after sales of neighbouring branches in at least 2 groups, per city.
In the Library
Open
Are the 18- and 30-month payback bands right when rent is paid a year in advance?
- Why it matters
- The payback bands decide whether a branch is funded; rent paid up front moves cash earlier than the model assumes.
- What would settle it
- Month-by-month cash of new branches under different rent schedules.
In the Library
Open
What share of aggregator orders is truly new, not moved from dine-in or takeaway?
- Why it matters
- A delivery order that only moved from the counter pays a commission and adds nothing.
- What would settle it
- A pre-registered holdout (the Channel Ledger's promotion test) in one group.
In the Library
Open
At what Geographic Replication Index does a new city fail?
- Why it matters
- The index measures how well the model travels, but no failure level has been observed yet.
- What would settle it
- GRI at months 6 and 12 for new-city branches, with their outcome at month 24.
In the Library
Open
How much of the Ramadan dip comes back in Shawwal, by branch format?
- Why it matters
- A branch judged in Ramadan can look weaker than it is; the recovery decides how to read it.
- What would settle it
- Weekly sales for 2 Ramadans across at least 3 formats.
In the Library
Open
Is the exit rule (two negative quarters) too slow for a small group?
- Why it matters
- A small group has less cash to carry a losing branch while it waits for the rule.
- What would settle it
- Cash position at the exit decision in groups of fewer than 10 branches.
In the Library
Open
After how many branches does head office grow faster than revenue?
- Why it matters
- Overhead that outgrows revenue eats the return on every new branch.
- What would settle it
- G&A and revenue growth by quarter, across group sizes (a question for the 2027 Index).
In the Library
Open
Does AI search change how Saudi customers choose a restaurant?
- Why it matters
- If customers ask an assistant before they visit, the signals a group owns change.
- What would settle it
- A repeated panel question, plus referral data from groups that track it.
In the Library
Open
Does the Branch Gate work outside restaurants and retail?
- Why it matters
- The gate's bands (18 and 30 months, 80% at peak) were written for F&B and retail branches. A sector with another capacity limit or contracted revenue, such as clinics, pharmacies or car services, may need other bands, or other tests.
- What would settle it
- Gate verdicts recorded before signing, compared with actual payback 12 months after opening, for branches in at least two sectors outside F&B and retail.
In the Library