Halima Mponda, treasurer of the women traders’ cooperative, began the meeting with a warning.
“The cooperative does not have one opinion.”
“I am asking for the contract, the relevant rebate rules, and the records needed to answer a consumer complaint.”
“And some members will say you are trying to kill the tour route.”
Mama Saada sat against the wall with her arms folded.
Halima placed the official contract on the table. Its version number matched the copy Rehema had seen at the stall.
Page seven was the same.
The appendix was more revealing.
Group codes were organized into categories connected to a **visibility rebate**, **guide conversion band**, and **weekly compliance factor**. Vendors following the categories received stronger route placement. Vendors who departed from them could lose points.
“Is there a clause saying the customer must see the group package before ordering?” Rehema asked.
Halima searched.
“There is a general rule that prices must be displayed.”
“But the group card is in a drawer.”
Mama Saada said, “The guide tells the group it is a package.”
“Verbal description is not the same as itemization.”
Halima produced rebate summaries but refused to photocopy entire statements because they contained individual vendor turnover.
“I do not need everyone’s revenue.”
Together they prepared an extract showing the rules and sample totals while redacting business information that was not needed.
One period showed Mama Saada leaving the scheme. Her rebate fell, but more importantly her route-placement score dropped.
“Does the route score prove the sales decline?” Rehema asked.
“No,” Halima said. “It is part of the sequence.”
Good.
A vendor named Fatuma entered and challenged the direction of the conversation.
“The scheme brought us tourists. You cannot speak as though everyone is a victim.”
Rehema asked permission to record that view.
Fatuma agreed.
She said guide traffic had increased sales and most customers never complained.
Rehema added the counterview to the file.
She did not want a report where every vendor became either a helpless victim or a scammer.
Halima described the association’s favorite language: vendors were free to choose prices, guides were free to choose stalls.
“But every freedom has a score,” she said.
“May I quote that exactly?”
“Yes, as treasurer.”
The rebate summaries carried batch IDs resembling ferry arrival codes.
One line read:
**F-112 / G-C1.**
Rehema did not treat a cooperative document as proof of what ferry data contained. She called Kassim and requested aggregate batch counts through the official consumer inquiry.
“No passenger names,” she said.
“Good,” Kassim replied. “And because the vendor is your mother, I am assigning an independent co-reviewer.”
Rehema agreed.
Before leaving, she asked Halima who could change the appendix. The association board issued framework updates; vendors signed renewals. Individual stalls could negotiate some details but not the central visibility bands.
That helped Rehema frame agency more accurately.
The vendors were not powerless. They could participate or refuse.
But refusing carried commercial consequences.
That distinction would matter later when they negotiated reform instead of claiming no one had chosen anything.
The investigation had moved from a single stall into the architecture that linked guide groups, vendor prices, rebates, and route visibility.
Halima also explained that the cooperative itself received pressure from divided members. High-volume stalls wanted the guide framework preserved; smaller stalls wanted route access without conversion targets. If Rehema published individual turnover figures, those internal divisions could become personal attacks. She therefore kept dependency data at category level. The restraint would later make it easier for vendors to negotiate together because the investigation had not exposed who was most financially vulnerable. Rehema began to understand that transparency about rules did not require transparency about every participant’s private business weakness.
The next question was whether the ferry batches actually matched those group categories in practice.