Education & SaaS

A tutoring network running 4 centres (Sydney, Melbourne, Brisbane, Canberra) delivering exam-prep and selective school coaching to 500–2000 students per year, testing students for Selective School Entrance (OC test, NAPLAN top 10%, Opportunity Class), HSC booster programmes (Years 11–12 extension maths, English, science), IB preparation (Year 12 international baccalaureate, maths AI/AA, English A, sciences), and university bridge courses (first-year STEM subjects). Tiers: Foundation (NAPLAN 40–60th percentile, grades 5–6, foundational maths/English), Standard (NAPLAN 60–85th percentile, grades 6–8, accelerated curriculum), Advanced (NAPLAN 85th+, grades 7–10, competition-level problem-solving, olympiad-style challenges). Revenue model: diagnostic test ($45/student, 2000 students = $90k), term-by-term enrolment (Foundation $120/month × 300 students, Standard $180/month × 150 students, Advanced $300/month × 50 students = $36k + $27k + $15k = $78k/month = $936k/yr), paper marking (tutors mark past exam papers, report included in term fee), parent reports (auto-generated, included in fee), and scholarships (30% discount to merit-based or low-income students, ~20% of enrolment = revenue loss $187k/yr, offset by student volume). Current system: diagnostic test = paper-based multiple-choice + short-answer administered in-centre or mailed to families, admin manually scores using answer key (2 hrs per 20 students = 100 students = 10 hrs/day × 20 days marking = 200 hrs/marking cycle, 1 marking cycle every 2 weeks in peak season = 2000 hrs/yr = $50k labour), scoring errors (tutor adds up maths score wrong, student scores 58th percentile instead of 68th, placed in Foundation instead of Standard, parent pays $120/month for 6 months for wrong tier = $720 loss + student frustration), no auto-tier recommendation (admin reviews scores, decides manually "this student is borderline Standard/Advanced, email them tier options, let them choose" = 50 borderline students per cycle × 26 cycles/yr = 1300 emails, uncertain tier placement = 10% enrol in wrong tier, fail mid-term, request transfer, admin re-assigns class and issues credit, processing 130 transfers/yr = 6.5 hrs admin labour per transfer = 845 hrs labour = $21k). Cohort assignment = manual email "Emma, you are Advanced tier. Enrol in ADVAN-ENG-JUN-WED-6PM." Parent misses email or hits "reply" instead of clicking enrolment link. Admin re-sends email. Parent still confused. Admin calls parent, verbally confirms tier, receives enrolment. Total admin touch per student = 15 min (resolve email/phone confusion) × 500 students = 125 hrs = $3.1k labour. Practice paper assignment = tutor email list with 10 past HSC papers attached. Papers are assigned manually per tier (no versioning, no tracking of which students got which papers). Tutor downloads, marks on paper, scans, emails results to admin. Admin receives 50 tutor emails, each with 5–10 scans (hundreds of PDFs), manually enters scores into spreadsheet ("Emma paper 1: 72%, paper 2: 68%..."). Data-entry errors (admin types "72" as "27", or marks duplicate, or loses PDF). Error rate = 3 errors per tutor per month × 50 tutors = 150 errors/month = 1800 errors/yr. 30% result in billing dispute or parent confusion = 540 disputes/yr × $50 labour to settle (email, explanation, re-submission) = $27k labour. Paper marking labour = system doesn't auto-aggregate scores, so tutors hand-mark 200 students × 10 papers per year per tier (Foundation: 300 students, Standard: 150, Advanced: 50 = 500 students × 10 papers = 5000 marks × 2 min per mark = 166 hrs marking per tutor, assume 1 dedicated marking tutor per centre = 4 tutors × 166 hrs = 664 hrs/yr = $26.6k marking labour at $40/hr specialist tutor rate). Tutors burnout (166 hours of hand-marking per tutor per year is unpaid curriculum prep, takes away from lesson planning). Progress reports = admin collects feedback from tutors (email: "send me Emma's progress for Jun report"), tutors email back loose notes, admin manually writes PDF report per student (2 hrs × 500 students = 1000 hrs/yr = $25k labour). Reports are inconsistent (some tutors detailed, some vague), late (some tutors forget to send feedback, reports delayed 2 weeks), and missing data (no attendance summary, no benchmark scores, just tutor opinion). Discount management = admin maintains "scholarship spreadsheet" with student names + discount % + enrolment date. Invoice system (separate) doesn't auto-apply discount. Admin manually enters scholarship note on each invoice before sending. 500 students × 12 invoices/yr = 6000 invoice lines. 5% are scholarships or bulk-enrolment = 300 lines requiring manual discount entry. Time: 10 sec per entry × 300 = 50 mins/month = 10 hrs/yr = $250 labour. But errors occur: admin forgets to apply discount to invoice, student sees $180 instead of $126, parent calls "why am I paying full price?", admin re-issues invoice, credits account, 20 errors/yr × $100 labour to resolve = $2k labour. Total bleed: $50k diagnostic marking labour + $21k borderline-tier labour + $3.1k cohort-assignment labour + $27k paper-entry-error labour + $26.6k paper-marking labour + $25k progress-report labour + $2k discount-error labour = $154.7k/yr + unquantified cost of student dissatisfaction (10% wrong-tier enrolments, student quits mid-term, lost revenue $120 × 6 months × 50 students = $36k/yr). Total bleed: ~$191k/yr labour + $36k revenue loss = $227k/yr.

Why Cluey, TutorCruncher & Kindling Fall Short for Exam-Prep Networks

Cluey (Australian online tutor platform, $50–150/month per student subscription) is 1-on-1 live tutoring, not group exam-prep cohorts. No diagnostic placement engine (Cluey matches student to 1-on-1 tutor based on subject + grade, not tier-based cohort), no practice paper workflow (Cluey tutors assign homework in-session, no paper-marking aggregation), no progress-report automation (reports are manual tutor feedback). TutorCruncher (UK-based, $129–399/month + per-student fees) has lesson scheduling and basic invoicing but no diagnostic testing module (you enter student level manually, no auto-tier recommendation), no paper-marking workflow (no system support for scanning, scoring, aggregating marks), no scholarship-discount rules (discounts are manual per-invoice, not automatic). Kindling (AU, $99–299/month + $3–5/student/month) has group lesson scheduling but no diagnostic engine, no paper-marking system, no auto-tier placement. All three lack the diagnostic-placement-to-tiered-cohort pipeline. Exam-prep networks need: (1) diagnostic test module (online intake test, auto-scored, tier recommendation), (2) cohort auto-assignment (student takes test, system recommends tier, one-click enrolment into cohort), (3) practice paper workflow (system assigns papers per tier, tutors mark in-system, marks auto-aggregate, analytics generated), (4) progress-report automation (system pulls marks + attendance + tutor feedback, generates PDF), (5) scholarship/discount rules (admin sets rules, system auto-applies at invoice). Cluey has none of these. TutorCruncher has basic scheduling but no diagnostic or paper workflow. Kindling has scheduling but no diagnostic, no marking system, no report automation. Result: exam-prep networks using generic SaaS spend $154.7k/yr manual labour (diagnostic marking, tier assignment, paper marking, report writing), suffer $36k/yr revenue loss (10% wrong-tier enrolments), and burn out tutors (166 hrs hand-marking per tutor per year unpaid). Custom platform handles all 5 steps automatically.

What Custom Replaces: Six Features Exam-Prep Networks Need

1. Diagnostic Test Engine with Auto-Tier Recommendation

May 15, 2026. Parent buys diagnostic test ($45 online) for 12-year-old Emma (wants to attempt Selective School entrance test in Year 7). Emma sits test online at home: 50 questions, maths + English reasoning (problem-solving, reading comprehension, grammar, numeracy, logic). Test is auto-scored in real-time. System calculates Emma's raw score (38/50 = 76%), percentile (76th), and recommends tier: "Emma scores at 76th percentile. RECOMMENDED: Standard tier (60–85th percentile). If Emma wants challenge, Advanced tier is achievable with focused prep on problem-solving." Parent sees recommendation on portal. Parent clicks "Enrol in Standard" (or "Enrol in Advanced" if they want stretch). Enrolment is instant—system confirms: "Emma is enrolled in ADVAN-MATHS-JUN-MON-4PM + ADVAN-ENG-JUN-WED-6PM. First class: Monday, June 17, 4pm. Tutor: Mr. Patel." Parent gets calendar invite. Tutor sees Emma auto-added to class roster. Zero admin touch. Borderline cases: if Emma scored 58/50 (59% = borderline Foundation/Standard), system flags for human review: "Emma is borderline. Recommend: take diagnostic again in 2 weeks to confirm, or tutor assessment during first class." Admin notifies parent: "Emma's diagnostic was close call. Recommend starting Standard tier with tutor check-in first lesson. If Standard is too hard, move to Foundation (no charge, first lesson free adjustment period)." This prevents wrong-tier enrolment and refund requests. Manual system (TutorCruncher / Kindling): parent submits application form. Admin reviews. Admin emails parent tier recommendation. Parent replies (or doesn't). Admin manually enrolls Emma. Process takes 5–7 days. Wrong-tier errors occur because admin guesses (no diagnostic). Borderline students enrol in wrong tier, complain mid-term. Custom system: diagnostic auto-scores, auto-recommends, parent enrolls instant, zero admin lag, zero wrong-tier guesses. Labour saved: 5 min per student × 2000 students = 166 hrs/yr. But more important: wrong-tier errors drop from 10% (200 students) to <1% (15–20 students due to borderline reassignments). Revenue loss prevented: 200 students × $120 (Foundation) × 6 months avg = $144k revenue loss eliminated (at $300/month Advanced tier, loss is $360k). Actually: conservatively assume 5% revenue retention (some families stick with wrong tier, some quit, refund credit issued). Still: $7.2k–18k revenue protected per cycle.

2. Auto-Cohort Assignment & Instant Enrolment

Diagnostic test is submitted May 15 at 8pm. System auto-scores, generates tier recommendation, sends email to parent: "Emma is Advanced tier (76th percentile). Your enrolment options: [BUTTON: Enrol ADVAN-ENG-JUN, start Jun 17] [BUTTON: Enrol ADVAN-MATHS-JUN, start Jun 17] [BUTTON: Enrol BOTH, save $60/month (dual-subject discount)]." Parent clicks "Enrol BOTH" at 8:23pm. System confirms: Emma is added to ADVAN-ENG-JUN (Wed 6pm, Tutor Sarah, Room 2) + ADVAN-MATHS-JUN (Mon 4pm, Tutor Patel, Room 3). Calendar invite sent to parent + student. Admin team sees notification: "1 new enrolment Advanced cohort (2 classes)." Zero admin action required to place Emma. Compare manual system (Kindling): admin receives test results. Admin email goes to parent: "Hi, Emma is advanced tier. Please reply with which classes you'd like." Parent receives email, busy week, replies Thursday. Admin manually logs into Kindling, navigates to Emma's profile, clicks "enrol ADVAN-ENG," clicks "enrol ADVAN-MATHS," generates invoice for 2 classes ($360), sends invoice email. Parent receives invoice, confused by format (looks like 1 invoice, but is it for 1 month or 2 classes?), calls admin. Admin re-explains. Parent pays. Total process: 5 business days, multiple emails, admin touch 3 times (email reply, system enrollment, invoice send). Custom system: 23 minutes. Parent receives email May 15 at 8:01pm (1 min after test submission), clicks buttons at 8:23pm, system auto-confirms, admin zero touch. Labour saved per student: 10 min admin time × 2000 students = 333 hrs/yr. At $30/hr admin cost = $10k labour saved. Plus: faster enrolment = faster revenue collection (instant invoice, not 5-day lag) = $936k/yr revenue collected 5 days earlier = $12.8k cash-flow benefit (assuming 5% discount rate = accelerated payment value).

3. Practice Paper Workflow with Auto-Mark Aggregation

Emma is enrolled in Advanced Maths. System auto-assigns 10 past HSC exam papers (2015–2024 papers, difficulty matched to Advanced tier). System generates "Paper Schedule": Paper 1 due Jun 23 (tutor will mark by Jun 28), Paper 2 due Jun 30, ..., Paper 10 due Aug 18. Emma receives email: "Your papers are ready. [LINK to Paper 1 PDF]. Complete by Jun 23, upload answer script [UPLOAD BUTTON]." Emma writes paper (2 hours), takes photo of answer script with phone, uploads to system. System stores photo + timestamp. Tutor Patel receives notification: "Emma uploaded Paper 1. [VIEW]." Patel downloads, marks in red pen (or digital markup), uploads marked PDF back to system. System extracts score (e.g., "74/100") and stores in Emma's record. System auto-emails Emma: "Paper 1 marked: 74/100 (74%). You scored above class average (71%). Feedback: [TUTOR'S TEXT: 'Great approach to Q5. Work on proofs in Q7—revisit module 3.']." Emma sees mark instantly, views feedback, studies tutor's notes. System aggregates Emma's marks: Paper 1 (74%), Paper 2 (78%), ..., Paper 10 (82%). Trend: improving. Average: 76%. System alerts tutor: "Emma's progress: 76% avg, +8% trend. Ready to move to competition-level papers?" Tutor can recommend Emma to stretch-group (if Emma wants extra challenge). Manual system (Kindling / TutorCruncher): admin emails Emma + 49 classmates 10 papers as PDF attachments. Students download, complete, email or hand in paper script to tutor. Tutor marks on paper, scans, emails PDF back to student + admin. Admin re-enters score into spreadsheet ("Emma paper 1: 74%"). If admin makes error, student sees wrong score. Admin sends email each Monday: "Class average: 71%. Here are your individual scores." Student doesn't get tutor feedback in writing. No trending. No automated alerts to tutor (tutor doesn't know Emma is improving and ready to stretch). Marking labour: 50 students × 10 papers = 500 mark entries. Admin time: 500 entries × 1 min per entry = 500 min = 8.3 hrs/cycle × 4 cycles/term = 33 hrs/term = 100 hrs/yr = $3k labour. Tutor time: scanning + email = 5 min per student = 250 min = 4 hrs/cycle × 4 = 16 hrs/yr. Custom system: auto-upload by student, tutor marks once (uploads marked PDF), system extracts score automatically (OCR, or manual entry in web form = 1 min per paper). Admin zero touch. Tutor time: 1 min per paper × 10 = 10 min per student per cycle = 5 hrs/cycle × 4 = 20 hrs/yr (similar to manual). But admin labour: 0 (system handles re-entry + storage + email). Labour saved: $3k/yr per centre × 4 centres = $12k/yr. Plus data quality: zero re-entry errors, automated trending, tutor insights (system auto-flags students ready to stretch, tutors don't miss opportunities).

4. Auto-Generated Progress Reports with Benchmarking

June 30, end-of-month. System auto-generates progress report for Emma (Advanced Maths cohort). Report includes: (1) Attendance: 4 lessons attended / 4 lessons scheduled (100%), (2) Paper marks: Paper 1 (74%), Paper 2 (78%), avg 76%, +8% trend vs first paper, (3) Tier benchmarks: Advanced tier avg is 73%, Emma is 76% (above tier average), percentile within tier: 68th (middle-upper range of Advanced students), (4) Tutor feedback: "[AUTOMATIC PULL from marks + any written notes by tutor during lessons] Emma is strong on algebra but needs practice on vector proofs. Assignment for next month: 5 additional vector worksheets. Recommendation: next term consider Olympiad-style extension track." (5) Next steps: "Emma is on track for HSC excellence. Recommend continuing Advanced tier through Year 12. Parent: book optional parent-tutor check-in to discuss stretch goals?" Report is 2 pages, PDF, auto-generated. Parent receives email + portal access. Parent reads report, sees Emma is performing well, feels informed. Manual system (Kindling): admin collects tutor feedback via email or verbal conversation. Admin writes PDF report manually (2 hours per student × 500 students = 1000 hrs/yr = $25k labour). Reports are inconsistent (some detailed, some vague), late (admin waits for tutors to send feedback, report delayed), missing data (no attendance pulled automatically, admin manually counts), and no benchmarking (report doesn't say "Emma is 76% which is above tier avg of 73%"—that requires admin to manually calculate tier stats, add to report). Parent reads report, has questions, calls admin. Admin doesn't have detailed data, refers to tutor. Tutor didn't write the report, doesn't remember all details. Frustration. Custom system: report generates automatically, pulls attendance (system has record), pulls marks (system auto-aggregates), calculates tier benchmarks in real-time (comparing Emma's 76% to cohort avg), includes tutor feedback (tutor can add free-text note "strong on algebra..." when marking papers or after lessons), report is consistent (template-driven), always on-time (auto-generated on-demand or on schedule), never late, data-rich. Parent gets clear picture. Labour saved: 1000 hrs/yr = $25k labour. Value-add: automated benchmarking, richer insights, higher parent satisfaction, higher retention (parents see progress clearly, stay enrolled longer).

5. Scholarship & Discount Rules Engine

Admin sets rules in custom system: (1) "Scholarship discount: –30% for merit-based (top 10% of cohort) or income-tested (eligible families). Mark student as 'scholarship' in profile." (2) "Bulk-family discount: –15% if 2+ siblings enrolled." (3) "Early-enrolment discount: –10% if enrol before May 1." System stores these rules. At invoice generation (e.g., June 1 invoice for Emma), system checks: Emma is in Advanced tier ($300/month base). Emma is marked "scholarship" in profile. System auto-applies rule 1: $300 × 30% = –$90. Discount is automatically shown on invoice: "Emma Advanced Maths Jun: $300 - Scholarship (–$90) = $210." Invoice sent. Zero admin touch. If Emma's parent enrolls her brother Liam in same cohort (both scholarship students), system checks: Liam is enrolled (rule 2 applies). Both students already have scholarship (–30%). System doesn't double-apply (–30% + –15% = –45% = system error). Instead, system shows: "Emma: $300 - Scholarship –30% = $210. Liam: $300 - Scholarship –30% - Bulk Sibling –10% (can't apply both; scholarship is primary) = $270 - Scholarship –30% = $189." System intelligently combines rules (no double-discount). Admin never touches invoice. Revenue is accurate. Manual system (Kindling): admin maintains spreadsheet with scholarship list. Invoice system doesn't know about discounts. Admin manually edits each invoice before sending: "Emma $210 (applied 30% scholarship discount)." 500 students × 12 months = 6000 invoices. 20% have discounts = 1200 invoices requiring manual edit. Time: 2 min per invoice (find student in spreadsheet, apply correct % mentally, edit invoice number) × 1200 = 2400 min = 40 hrs/month = 480 hrs/yr = $14.4k labour. Errors: 5% of manual discounts are wrong (admin applies 15% instead of 30%, or forgets discount, or applies to wrong month). 60 errors/month × 12 months = 720 errors/yr. Cost to resolve: email to parent, re-issue invoice, credit account = $50 per error = $36k labour + revenue loss (if error is "we charged full price instead of discounted," we're owed the difference but wait 2–3 weeks to get paid = cash-flow hit). Custom system: 0 labour (rules auto-applied), 0 errors (rule engine is deterministic, no manual entry), revenue accurate, cash-flow smooth. Labour saved: $14.4k/yr. Error/revenue loss prevented: $36k/yr. Total: $50.4k/yr saved.

6. Recurring Billing & Payment Reconciliation Automation

Emma is enrolled in Advanced Maths ($300/month, cohort fee) + Advanced English ($200/month, cohort fee) = $500/month. June 1 invoice generated automatically: "Emma Term 1 (Jun–Jul): Advanced Maths Jun $300 + Advanced English Jun $200 = $500. Scholarship –$90. Total due: $410. Due date: Jun 10. [PAY NOW button]." Invoice is sent to parent. Parent clicks "Pay Now," pays via Stripe (credit card). Payment is instant, system reconciles: "Invoice marked paid, $410 received, $0 outstanding." July 1 invoice auto-generated again: "$500 - $90 scholarship = $410. Due: Jul 10." Same process. System tracks: June $410 paid, July $410 paid, August $410 paid, ... If parent misses payment (July invoice unpaid as of Jul 15), system auto-sends reminder: "Emma's lesson payment is 5 days overdue ($410). [PAY NOW button]." Parent pays on Jul 17 (2 days late). System records: "Paid 2 days late, automatically sent 1 reminder." Admin dashboard shows: "July reconciliation: 500 students invoiced for $500 each = $250k. Received $247.5k (99% collected). Outstanding (7+ days overdue): $2.5k (5 families). Sent reminders to 5 families. Expected payment within 5 days." Zero manual reconciliation. Manual system (Kindling): invoices are generated monthly. Admin exports CSV from Kindling, pastes into Excel, manually tracks payments (received vs outstanding). Payments come via bank transfer (some families) + Stripe (some families). Admin must check bank statement, match to invoice (invoice says "Emma $410" but payment memo is vague "E.L. payment"). 10% of payments don't match automatically = 50 unmatched payments/month = admin spends 10 min per unmatched payment (email to family "we received $410 on Jul 2, is this for Emma June?", family replies, admin manually matches) = 500 min = 8.3 hrs/month = 100 hrs/yr = $3k labour. Collections: if payment is overdue, admin sends email reminder. If still unpaid after 2 reminders, admin calls family. Some families pay, some have excuses ("money is tight, can we pay half this month?"), admin negotiates, manually adjusts invoice. Manual overrides = 10 families/month × 12 = 120 manual overrides/yr. Admin time: 15 min per override = 1800 min = 30 hrs/yr = $900 labour. Unresolved overdue: 5% of invoices go past 60 days unpaid = 30 invoices/month × 12 months = 360 unpaid invoices/yr × $500 avg = $180k unpaid revenue (eventually collected, but 60–90 day lag = cash-flow hit). Custom system: payment reconciliation is automatic (Stripe webhooks auto-match, bank transfers auto-matched via reference code), reminders are auto-sent, collections are pro-active (system flags overdue, admin gets alert, admin can manually intervene if needed). Labour saved: $3.9k/yr labour (matching + reminders). Cash-flow improved: $180k outstanding is reduced (system reminds proactively, reduces 60-day lag to 14-day lag) = accelerated payment = $2.7k cash benefit (assuming 5% discount rate, 46-day acceleration).

Australian Exam-Prep Regulatory Context

Selective school exam-prep in Australia involves: Selective School Entrance (NSW, VIC, QLD: NAPLAN-based cutoff for entry into selective public schools like James Ruse, Melbourne High, Brisbane Grammar, requires NAPLAN score 85th+ percentile or higher), Opportunity Class (NSW: Year 5 entrance into gifted cohorts, requires NAPLAN 90th+ percentile + problem-solving test), Overseas admission (IB, VCE, HSC: international schools require Australian evidence of academic strength for scholarships, custom testing by school), scholarship exams (private schools: entrance exams for merit scholarships, problem-solving + reasoning), and NAPLAN booster courses (coaching students to improve NAPLAN scores in Year 3/5/7/9 before national test). Regulatory: no state-specific licensing for exam-prep tutoring (unlike childcare or schools, tutoring centres are not regulated), but duty of care applies (centre responsible for student wellbeing, safe environment), WWCC checks required if centre employs staff (tutors, admin) who work with children, and consumer law applies (Australian Consumer Law: tutoring contracts must have clear terms, refund policy, cancellation terms, no misleading advertising of results like "95% of our students score 90th+ percentile" without evidence). Custom system should track: exam-type per cohort (Foundation is NAPLAN 40–60th prep, Standard is 60–85th, Advanced is 85th+ / Selective School / OC prep), student age + year level (cohorts are year-level grouped: Year 5 NAPLAN, Year 7 Selective Entrance, Year 11 HSC), results tracking (aggregate marks per cohort, anonymized for reporting: "Foundation Maths avg 65%, Standard avg 78%, Advanced avg 84%," never "Student X scored 40%"), and duty of care compliance (WWCC checks for staff, incident log if student wellbeing issue arises, safe environment policies). Custom system doesn't replace policies, but enables compliance (track WWCC, store policies in student welcome pack, log any safety incidents). Terminology: system should use exam-aware language (Foundation / Standard / Advanced are tier names, but underlying is NAPLAN percentile brackets + exam-readiness). System should NOT make promises like "our tutoring guarantees 90th percentile NAPLAN score" (not evidence-based, violates ACCC truth in advertising). System SHOULD report: "Emma was 74th percentile in diagnostic, placed in Standard tier, current papers average 76%, which is above Standard tier average (73%)." Factual, evidence-backed, transparent.

Six FAQs

Can we start with 1 centre and scale to 8 later?

Yes. Custom system is multi-centre from day one. You configure Centre A (Sydney) with cohorts (Foundation Maths, Standard Maths, Advanced Maths, etc.). When you open Centre B (Melbourne), you create Centre B in system, replicate cohorts (system copies cohort templates, you adjust tutor names + schedule for Centre B). Diagnostic test scores are global (a student in Sydney or Melbourne takes same diagnostic, auto-tiered to same tier structure). Enrolment tracks per centre (Emma enrolls in Sydney Advanced Maths, her brother in Melbourne Standard Maths, system keeps them separate but linked for bulk-sibling discount). No per-centre licensing fees. Scale from 1 → 8 centres, system cost stays the same ($120–180k upfront build, $3k/yr hosting). Compare: Kindling charges per centre + per student. 1 centre (300 students) = $99–299/month per centre + $3–5/student/month × 300 = $150/month + $900–1500/month = $1050–1650/month = $12.6k–19.8k/yr. 8 centres (2400 students) = $800/month base + $7200–12k student fees = $8000–12.8k/month = $96k–153.6k/yr. Per-seat fees creep with scale. Custom system = no creep.

What if a tutor works across 2 centres (Sydney + Melbourne)?

Tutor profile is global. Mr. Patel teaches Advanced Maths at Sydney Centre (Mon + Tue, 4–6pm) and Melbourne Centre (Thu + Fri, 4–6pm). Rosters are per-centre (Sydney shows Patel Mon–Tue, Melbourne shows Patel Thu–Fri). System prevents double-booking: if admin tries to add Patel to Sydney Wed 5–6pm and Melbourne Wed 5–6pm (time overlap), system alerts "Patel is busy Wed 5–6pm Sydney. Cannot add Wed 5–6pm Melbourne (same time zone, 30 min travel time required)." Cross-centre consistency: Patel's WWCC expiry date is stored once (global). If Patel's WWCC expires Aug 15, both centres see alert: "Patel unavailable after Aug 15 (WWCC expired)." Salary/hours: admin tracks Patel's total hours across both centres (Mon–Fri, 4 hrs/week = 16 hrs/week, $25/hr = $400/week). System can generate payroll per centre (Sydney: 8 hrs/week = $200, Melbourne: 8 hrs/week = $200) or global (16 hrs/week = $400). Multi-timezone handling: if you expand internationally (Sydney + London centre), system handles time zones. Tutor can't work simultaneous lessons across zones (system auto-prevents).

Can diagnostic test be done remotely?

Yes. Diagnostic is online, delivered via link. Student can take test at home (on any device: desktop, laptop, iPad) or in-centre (proctored by admin). Online test = student logs in, clicks "Start Test," answers 50 questions (multiple-choice + short-answer), system auto-scores, result is instant. Remote proctoring (optional): if centre wants anti-cheating measures, admin can enable "invigilation mode" (student shares screen, admin watches, student cannot open other tabs). Most tutoring networks don't require proctoring for intake diagnostic (students aren't motivated to cheat on diagnostic; they want accurate placement). Proctoring is only needed for official exams (NAPLAN, Selective Entrance, HSC). For intake diagnostic, remote unproctored is fine. System logs: student name, test date, IP address, device, raw score, percentile, tier. If student retakes test 2 weeks later (to confirm tier), system compares: "Emma's test 1: 76th percentile, test 2: 78th percentile. Recommendation: Advanced tier confirmed (both tests 75th+)."

How do we handle makeup lessons or cohort changes?

Makeup lesson: if Emma misses Standard Maths (scheduled Wed 6pm, she was sick), admin can reschedule to a different timeslot in same tier (same tutor, same lesson plan, if tutor has capacity). System tracks: "Emma missed Wed Jun 13 (illness, excused). Rescheduled to Thu Jun 14, 5pm (tutor available, same lesson plan)." Attendance still logs at rescheduled time. Invoice doesn't change (1 lesson $200, paid whether Wed or Thu). Cohort change: if Emma's paper marks show she's struggling in Advanced (avg 62% vs cohort avg 78%), tutor recommends: "Emma should move to Standard tier to build confidence." Admin approves change. System moves Emma from Advanced Maths to Standard Maths (next month). Invoice adjusts: next month Emma pays Standard rate ($180/month instead of $300). System pro-rates current month if change is mid-term: if Emma was in Advanced June (paid $300 for full month), and moves to Standard July, system generates credit: "June overpaid: Advanced $300 – Standard $180 = $120 credit. Applied to July invoice: $180 - $120 = $60 due." Automatic, zero admin touch. Alternate: if Emma wants to move up (struggling Advanced student wants to stay for challenge), system allows: tutor provides extra support, system monitors marks, if Emma improves, stays. If Emma continues to struggle after 4 weeks, system alerts tutor: "Emma is still averaging 65% in Advanced (below cohort 78%). Recommend Standard tier reassignment?" Tutor can recommend, or Emma can choose to persist.

Can we offer refunds if a student is misplaced in wrong tier?

Yes, built-in. Refund policy: "If student is misplaced in wrong tier and wants to change within 2 weeks of starting, one of the following: (a) move to correct tier (pro-rata invoice adjustment, no refund), or (b) withdraw and refund balance (if paid in advance, refund unused portion). After 2 weeks, no refund, tier change allowed with pro-rata adjustment." System enforces this: if Emma enrolled Jun 1, paid $300 for June Advanced Maths, and requests refund on Jun 5 (within 2-week window), system can auto-process: "Refund requested Jun 5 (day 4 of 28-day month). Days remaining: 24. Pro-rata refund: $300 × (24/28) = $257. Refund $257 to original payment method (Stripe)." Or Emma can move to Standard (pay Standard rate $180 for June, move forward). System tracks refund reason (misplaced tier, personal reason, changed mind, etc.) for reporting. Most misplacements are caught within 1 week (Emma takes first lesson, tutor assesses, says "you're too advanced/too basic," admin intervenes, moves Emma, zero refund needed). Strong diagnostic + tutor check-in = <1% refunds.

How do we prevent tier gaming (parents pushing kids into Advanced to appear smarter)?

Tier placement is diagnostic-based, not parent choice. Diagnostic is auto-scored (objective, can't game). Parent can see recommendation, but can't override diagnostic-recommended tier. If parent insists on Advanced despite Foundation diagnostic result, system allows: "Emma is Foundation tier (54th percentile) but parent requests Advanced tier. [PROCEED] [BACK]. Warning: Emma may struggle with Advanced curriculum. System will monitor Emma's paper marks (target 75%+). If Emma averages <60% after 4 weeks, system will alert tutor to recommend tier adjustment." Parent clicks "Proceed at own risk." System tracks override (admin can review: "15 parent overrides this month, 5 resulted in mid-month tier changes due to struggle"). This prevents abuse (if parent pushes kid into Advanced, kid struggles, parent requests refund 2 weeks later, system says "Emma was correctly placed in Foundation, override was parent choice, no refund"). Tutor also has power: tutor can recommend tier change anytime if student is clearly misaligned (e.g., "Emma is Foundation but is acing every paper and asking advanced questions, recommend move to Standard after 2 weeks").

The Bottom Line

Exam-prep tutoring networks testing 500–2000 students per year need diagnostic placement tests, auto-tiered cohort assignment, practice paper workflows with aggregated marking analytics, auto-generated parent reports with benchmarking, and scholarship-discount rules. Cluey (1-on-1 tutoring platform), TutorCruncher (generic lesson scheduler), and Kindling (group lesson platform) all lack diagnostic engines and paper-marking workflows. Manual diagnostic marking costs $50k/yr labour. Manual cohort assignment costs $3.1k labour + $21k borderline-tier labour. Manual paper marking + data entry costs $27k + $26.6k labour. Manual progress reports cost $25k labour. Manual discount application costs $14.4k labour + $36k error/revenue loss. Total bleed: ~$227k/yr labour + $36k revenue loss + tutor burnout (166 hrs hand-marking per tutor per year unpaid). Custom platform automates all 6 steps: diagnostic auto-scores + recommends, enrolment is one-click (parent clicks "confirm tier" button, Emma is enrolled, zero admin), paper marking is uploaded by students, scored by tutors in-system, marks auto-aggregated, analytics auto-generated, progress reports auto-generated (pulling marks + attendance + benchmarks + tutor feedback), discounts auto-applied at invoice. Build cost: $120–180k (diagnostic engine + cohort auto-assign + paper marking workflow + report generation + discount rules). Year one: $150k investment. Year two: $3k/yr hosting. Break-even: month 20. Year 2 onwards: $227k/yr labour saved + $36k revenue loss prevented = $263k/yr net (at 500 students, $526/student). Scaling to 8 centres × 500 = 4000 students: $2.1M/yr value. Generic SaaS per-student fees creep with scale; custom scales infinitely. Own your diagnostic placement. Own your cohort assignments. Own your paper workflows. Own your progress reporting. Own your discounting. Build custom. Ship faster. Scale to 8 centres without per-seat licensing bleed.

Ready to build a custom exam-prep platform for your tutoring network? Check Aidxn's custom software packages, or book a call to discuss your current student volume (500–2000/yr?), centre count (1–8?), exam types (OC, NAPLAN, HSC, IB?), cohort tiers (Foundation, Standard, Advanced?), paper workflow (how many papers per student per term?), and scaling goals (grow to 10+ centres and 10k+ students).