Upinion · Web App 2.0 · Day 2 — Istanbul · 27 May 2026

Four decisions to lock today. Then we ship.

Yesterday closed with a commit to WhatsApp + a 25-page interactive prototype. Today is the day to convert that into a concrete buildable plan — by walking out with stack, V1 flows, partner strategy, and pilot scope all locked.

Audience: Anvilda Istanbul team-days · Day 2 · Tracker: co-upinion#42

✓ LockedWhat Day 1 settled — please don't re-litigate

○ OpenThe four decisions for today

1

Stack pick — Turn.io vs Glific

This is the unblocker. Everything downstream (V1 scope, pilot timeline, what we promise the Embassy) depends on this. If we don't land it before lunch, the rest of Day 2 stalls.

Turn.io
Hosted SaaS (Public Benefit Corp)
~$120/mo with 40% NGO discount

Pros

  • Native WhatsApp Flows support — ready today
  • Zero engineering setup
  • Production track record: WHO COVID (14.7M), CARE Bangladesh (34M), MomConnect, Taskeen
  • Free 3-month Chat for Impact Accelerator + up to $25K possible support
  • Compliance-friendly (Cape Town based, GDPR-aware)

Cons

  • Proprietary — vendor lock-in
  • Data on Turn's servers (Cape Town) — sovereignty trade-off
  • Monthly fee scales with msg volume
Glific
Open source (AGPL-3.0)
~$115/mo hosted, free if self-hosted

Pros

  • Full data sovereignty — self-host on Upinion infra
  • Project Tech4Dev, India — 125 orgs, 100M+ msgs, DGPA-recognised
  • Open source — no vendor risk
  • Active maintainer + community

Cons

  • WhatsApp Flows NOT yet supported (planned post-Gupshup migration, no firm date)
  • Without Flows, we fall back to chat-only branching — Architecture B from the prototype
  • Self-host = ops burden (Upinion does not currently run servers)
Upii's recommendation — team owns the commit

Turn.io for V1. Flows support is the dealbreaker — Glific can't deliver the Round-3 prototype patterns today. Revisit Glific in 6 months when their Flows support ships. The $120/mo is within any reasonable pilot budget; the NGO accelerator gives us optionality.

30 min — decide before 10:30

2

Production V1 — pick the flows

All 4 prototype flows (A/B/C/D) are theoretically buildable. V1 should ship the 2 that have highest data-value × lowest partner-dependence. The other 2 wait for V1.5 / V2.

Flow B — Upinion asks (survey)
Core data collection
★★★ Highest value ✓ Ready

Notes

  • Direct fulfilment of the NL Embassy data-collection mandate
  • Topup mechanic motivates completion
  • Reloadly integration is the only external dep
  • Zero partner-org dependency to launch
Flow A — User reports info up
Bi-directional intel
★★★ Highest value ✓ Ready

Notes

  • Captures route changes, border-control incidents, smuggler activity — the qualitative gold
  • Augustine + Tarek both flagged the report button as a keeper from R1
  • Photo upload + Flow form are native; voice-note workaround established
  • Zero partner-org dependency
Flow C — Info inside a Flow
Outbound info delivery
★★ High value △ Partial — needs content pipeline

Notes

  • News-in-Flow + Channel broadcast = the value-give-back the team wants
  • Editorial overhead: someone has to write + verify weekly content
  • Best paired with a Croix Rouge / IFRC / IOM content partnership (not blocker but better)
Flow D — Free 1-on-1 with human
Human routing layer
★★ High value ✗ Needs ops rotation

Notes

  • Requires 2–3 humans on rotation for 4h SLA — Upinion does not currently have this
  • Highest emotional value (responds to vulnerability) but highest ops burden
  • Defer to V1.5 unless we partner with a service provider (IRC field office, MSF, etc.)
Upii's recommendation — team owns the commit

Ship B + A in V1. Both are zero-partner-dependency, high-value, and the prototype's strongest patterns. C in V1.5 (requires content pipeline). D in V2 (requires ops rotation OR partner).

20 min — decide by 11:00

3

Partner orgs for warm referrals (trust at first contact)

Without partner endorsement, ‘Anwar' is just another suspicious WhatsApp link — Lebanon hacking precedent says this kills onboarding. Each partner adds trust + the right partner unlocks geographic reach.

Croix Rouge Algérienne (CRA)
In-country, principled neutrality
Existing relationship: Likely — Upinion has CR contacts elsewhere
Value-add: Strongest trust signal in Algeria. CRA presence in Tamanrasset means they touch the migrant population physically. Can share number via flyers, in-shelter posters.
Ask: Distribute the WABA number through their network. No financial ask.
MMC (Mixed Migration Centre)
Research consortium
Existing relationship: Yes — we cite their 4Mi data extensively
Value-add: Adds methodological credibility. Their corridor reports could embed our Channel link. Already trusted by migrants reading those reports.
Ask: Co-publish a corridor brief + endorsement of methodology.
IRC field office (Niger / Mali)
Humanitarian frontline
Existing relationship: Indirect — Signpost (IRC) is the WhatsApp-NGO precedent
Value-add: IRC handles the post-expulsion side (Assamaka, Agadez). Migrants returning from Algeria pass through them — a referral here closes the loop.
Ask: Distribute the Anwar contact at Niger reception points + cross-referral protocol.
IOM Algeria
Inter-governmental
Existing relationship: Unknown — diplomatic-channel via NL Embassy
Value-add: Currently the only large humanitarian presence in Algeria with migrant access. Trust-amplifier if they vouch.
Ask: Co-publish a hotline directory in our Flow C content. May require Embassy-level conversation first.
Upii's recommendation — team owns the commit

Lead with CRA + MMC — both are likely-yeses with lowest political friction. IRC is a fast-follow once we have Niger/Mali corridor traction. IOM is highest-value but slowest — push it through the Embassy track, don't block V1 on it.

20 min — decide owners + timelines by 11:30

4

Pilot scope for NL Embassy proposal

We need a number to put in front of the Embassy. Too small = not worth funding. Too big = unrealistic without partners in place. Round-2 research gives us the corridor maths; we need a commitment.

Single-city pilot — Tamanrasset
Geographically narrow
200 active users · 12 weeks · 1 corridor
~€8K (topups €3K · Turn.io €0.4K · partner-coord €1.5K · ops €3K)
Upside: Tight, measurable, defensible. Tamanrasset is THE migrant hub. CRA partnership unlocks it.
Downside: Embassy may want broader corridor coverage to justify funding.
Two-city pilot — Tamanrasset + Adrar
Mid-scope
500 active users · 16 weeks · 2 corridors
~€20K (topups €7.5K · stack €0.6K · partners €4K · ops €8K)
Upside: Tests transferability across cities. Adrar = the central-Sahel route junction. Stronger story for Embassy.
Downside: Doubles partner-coord overhead. CRA + a second-city partner needed.
Full-corridor pilot
Wide
1,500 active users · 24 weeks · Algeria + cross-border to Niger
~€55K
Upside: Embassy-grade headline number. Cross-border data is the unique value.
Downside: Requires IRC Niger + IOM Algeria + 3+ Algeria cities. Too many dependencies for a pilot that hasn't yet proven the tech.
Upii's recommendation — team owns the commit

Two-city pilot (Tamanrasset + Adrar). Tight enough to ship cleanly, wide enough to be a real story for the Embassy. €20K is a credible ask. Pre-pitch CRA + a second-city partner BEFORE writing the proposal — verbal alignment first, paper second.

25 min — decide by 12:00

✓ Walk-outWhat we need committed before we leave the room

End-of-day checklist

⊙ ParkedTracked for after Day 2 — not on today's agenda

Quick links

▶ Interactive WhatsApp Flows prototype
The 25-page click-through from last night
🧭 Round-3 hub
Full research + audit trail
⚖️ Fork in the road — DECIDED
Why we picked WhatsApp