Kassim alimteua Pendo Mushi, officer wa visitor services kutoka shift nyingine, kuwa co-reviewer.
“Rehema ataendelea kama intake lead,” alisema. “Lakini conclusions kuhusu route/pricing zitasainiwa na wote wawili.”
Rehema alipoteza sole control ya case.
Alijisikia relieved na irritated kwa wakati mmoja.
Pendo akasema, “Hiyo ndiyo point.”
Walifungua aggregate ferry batch dashboard. Hakuna majina ya abiria, passports, phone numbers wala bookings binafsi. Waliona tu batch IDs, arrival times, group counts na tour-coach assignment categories.
F-112 ilikuwa batch ya feri ya saa tano.
Guide category **G-C1** ilikuwa kwenye subset ya groups.
Rehema alilinganisha risiti voluntarily submitted na batch times. Pendo ndiye aliyesimamia spreadsheet ili Rehema asichague data inayounga mama yake au kumhukumu.
“Tuna complaints ngapi with receipts?” Pendo akauliza.
“Nane mpaka sasa.”
“Small.”
“Ndiyo.”
“Basi no population claim.”
Walitenganisha timestamps.
Same stall, same dish, same wholesale day, tofauti ya guide code.
Lakini Pendo aliona possibility nyingine.
“Group size inaweza ku-change service.”
Rehema akakubali.
Waliongeza columns: itemized package? group service stated? guide code? demand period? wholesale surcharge?
Kwa risiti nyingi, higher price haikuwa na itemization.
Kassim akauliza, “Una proof route codes zinaamua price?”
“Association contract inazi-link kwa pricing categories. Ferry batch analysis inaonyesha groups zenye codes hizo zinafika same time. Hatujathibitisha every stall follows.”
“Good.”
Kabla shift kuisha, complaint nyingine ikaingia.
Then another.
Stalls nne tofauti, si Saada Seafood pekee.
Pendo akasema, “Now we need field verification.”
“Consent, no undercover.”
“Agreed.”
Rehema aliandika disclosure update: conflict with mother’s stall remains; scope expanded to market practice; independent co-reviewer assigned.
Kassim akasaini.
That step mattered because Mama Saada’s stall was no longer only evidence. It was one participant in a wider system.
Rehema alipotoka office, mama yake alikuwa amemtumia message:
*Wafanyabiashara wanasema umepeleka ferry records kwa association.*
Rehema akajibu:
*Aggregate batches only. No passenger data. Co-reviewer amepewa.*
Mama Saada hakujibu.
Dakika tano baadaye, Pendo alipokea complaint ya stall nyingine yenye risiti mbili na stamps **G-A2** na **WALK-IN**.
Pendo akaweka kwenye queue.
“Kesho market survey.”
Rehema alitazama complaints count.
Nane zilikuwa zimekuwa kumi na mbili.
Pendo aliingia review akiwa na suspicion tofauti.
“Maybe tourists wenye guides wanapata service nyingi zaidi na ndiyo maana complaints zinaonekana price tiers.”
“Possible.”
“Maybe ferry batches zinakuja peak hours na demand ndiyo price.”
“Possible.”
Walitengeneza hypotheses mbili ambazo zingevunja case rahisi ya discriminatory pricing.
Aggregate data ilionyesha guide batches na walk-ins zilifika overlapping times, hivyo peak demand alone haikueleza differences zote.
Lakini sample ilikuwa ndogo.
Pendo akasema, “Need more stalls.”
Rehema akaenda visitor desk logs, si passenger system. Complaints zote ambazo visitors voluntarily submitted zingeweza kuingia, lakini only price/receipt details.
Kassim akafanya privacy check.
“No copying tour bookings.”
“Agreed.”
Waliona batch code mechanics. Some group codes zilirudia across stalls.
Pendo akasema, “This shows operational link, not customer targeting by nationality or person.”
“Exactly.”
Rehema alitaka kuwasiliana na guide coordinators directly.
Pendo akamzuia.
“First vendor survey. Otherwise guide answers inaweza ku-shape questions.”
Rehema akakubali.
Co-reviewer independence ilikuwa inafanya kazi kwa sababu Pendo alikuwa willing kumpinga.
Kassim aka-review conflict note na kuongeza rule: any recommendation affecting Saada Seafood specifically must be signed by Pendo or another independent reviewer.
“Na general market recommendations?” Rehema akauliza.
“Joint.”
Hiyo ilipunguza power yake zaidi.
Mama Saada aliposikia, alimtumia: *At least now sio wewe peke yako.*
Rehema alijibu: *Ndiyo.*
Wakati complaints nne mpya zilipoingia, mmoja ulikuwa from vendor mwenyewe.
Vendor alisema customer amemuita tapeli kwa sababu blog imeanza kusambaza screenshot ya receipt.
Rehema hakutaka investigation iongeze harassment.
Aliweka notice kwenye visitor desk: complaints welcome; do not post vendor personal details; price dispute is under review.
Pendo akasema, “Good. Consumer protection si mob protection.”
They scheduled open market interviews na posted-price comparison.
No hidden cameras.
No sending fake tourists.
They also prepared category definitions kabla field: - seasonal/input surcharge, - volume discount, - optional service package, - guide-linked unexplained tier.
Rehema aliona kwamba evidence discipline inaweza kufanya story less dramatic but more actionable.
Zamu ilipoisha, four-stall complaints zilikuwa enough kuondoa focus kutoka mama yake pekee.
Pendo akasema, “Tomorrow we find whether Saada is exception or system.”
Rehema akajibu, “Or both.”
At the end of the batch analysis, Kassim required Pendo and Rehema to write a reproducibility note. Another officer should be able to take the same aggregate batch table and the same consented receipts and reproduce their comparisons. They saved formulas, not passenger data. Pendo deliberately swapped the order of several rows and reran the checks to make sure the pattern did not depend on sorting. It held. Rehema liked the exercise because it turned a suspicion about ferry codes into a process another employee could challenge after she left. That was credibility she could not create through confidence alone.
Pendo alipomaliza spreadsheet, ali-export summary yenye rows za batch na receipt categories tu. Kassim aka-run spot check ya formula dhidi ya source counts. Rehema hakushika raw passenger system. Hiyo separation ilifanya finding iwe audit-able bila kuunda database mpya ya tourists. Pendo pia aliandika cases ambazo haziku-fit pattern—risiti mbili zenye guide codes lakini prices sawa. Rehema alikubali ziwe kwenye report. Pattern yenye exceptions ni bora kuliko story inayoficha exceptions. Field survey ya kesho ingejaribu kueleza kwa nini baadhi vendors walifuata code bila kubadilisha bill.
Na sasa practices za stalls nne tofauti zilikuwa kwenye record.