|
Prepared for
Samer Alsousi — Director of Marketing
Mohammad Al Tallini — Sales Development Manager
Prepared by
Core88
Shareef Nuseibeh — Founder & CEO
Date
July 2026
1
Background
Following our presentation to the Teeba Sales & Trade Marketing team, there was clear interest in an image-recognition (IR) system that turns a single shelf photo into objective reporting on On-Shelf Availability (OSA), Share of Shelf (SOS), and availability per store. Before committing to a full rollout, Teeba and Core88 are collaborating to run a focused, paid two-week pilot to validate the model on Teeba’s own products and shelves — proving that the system works, improves reporting efficiency and market insight, and is feasible on cost.
2
Objective
Prove, on real Teeba shelves, that the IR system:
1
Reads OSA, SOS, and store availability accurately enough to be trusted for decisions;
2
Produces these insights faster and more consistently than today’s manual process;
3
Does so at a cost and operating model that scale to a full rollout.
=
Success means meeting the agreed SLA (Section 7). If met, both parties align on a full rollout engagement.
3
Scope
In scope
✓One Teeba product category (selected by the Teeba team).
✓10 retail accounts (selected by the Teeba team).
✓Manual baseline: Sample of current audit/report plus time and frequency.
✓Deliverables: a PDF report plus a live dashboard demonstration.
Out of scope
Reserved for later phases
×Planogram compliance scoring.
×Large-row / multi-bay image stitching (panorama). Each photo is analyzed as a standalone frame.
×Competitor SKU-level identification. Non-Teeba facings are counted as “Unknown” — enough to compute SOS without naming competitor brands.
4
Approach & Timeline
The pilot runs over two weeks.
Week 1
Intelligence & model prep
Core88 collects the category inputs, trains and extends the recognition model on Teeba’s SKUs, and builds the reporting template. From Teeba: 10 category images per store and the MSL per account.
Week 2
In-market validation
A joint Core88 + Teeba in-person store run captures real working-store conditions. In parallel, existing cloud images are run against the model. Outputs are validated against manual ground truth to measure SLA performance.
Note
The Week 1 clock starts once Teeba delivers the category inputs: images, MSL, and SKU master.
5
What We Need From Teeba
Capture protocol
One bay / section per photo, straight-on, in focus, full shelf visible.
6
Deliverables
Per-store report PDF
For each account: a SKU table (facings · SOS% · OSA vs. MSL · status), annotated photo evidence, and a prioritized gap / action list. Aggregate summary PDF
OSA and SOS across all 10 stores, an availability view by store, top gaps, and a model-vs-manual validation table demonstrating SLA performance. Live dashboard demo
A walkthrough of the Core88 dashboard showing the same insights in an interactive view — a preview of the rollout experience.
7
Service Level Agreement
Acceptance threshold: metrics met across ≥ 8 of 10 stores. Recognition accuracy
Overall recognition accuracy (blended)≥ 90%
SKU detection rate≥ 90%
SKU identification accuracy≥ 90%
False-positive rate≤ 10%
Business-metric accuracy vs. ground truth
OSA accuracy (against MSL)≥ 95%
MSL compliance accuracy≥ 95%
Share of Shelf± 5 pts
Facing count per SKU± 10%
Operational
Report turnaround from image receipt≤ 4 hrs
Correction / re-run turnaround≤ 48 hrs
Store & in-scope SKU coverage100%
Efficiency
Reporting time reduced vs. a manual audit of the same store
≥ 70%faster
Teeba input
Image usability ≥ 90%, photos meeting the capture protocol. Sub-standard images are flagged and excluded, not counted against model accuracy.
8
Roles & Responsibilities
Core88
Model training
Image processing
Reporting
Dashboard
Validation
Joint store run
 Teeba
Category & account selection
SKU master
MSL
Week 1 images
Manual baseline
Merchandiser participation (Week 2)
9
Commercials
Pilot fee
1250
JOD
2-week engagement
Payment in full upon kickoff
A discounted validation rate, limited to one category and covering OSA, SOS, and availability only. It is positioned to cover Core88’s cost to prove the system on Teeba’s shelves. Pricing at full rollout scales with the number of categories, accounts, and added capabilities — planogram, stitching, competitor identification.
10
Path to Rollout
On meeting the SLA, Core88 and Teeba move to a rollout engagement.
1
Align
Goals, scope, KPIs
2
Kickoff
Plan, data access, timeline
3
Implement
Train on full SKU set, set up reporting
4
Phased rollout
By region & category, milestone-based
Agreed & accepted
For Teeba
Name, title · Signature · Date
Shareef Bilal Nuseibeh
For Core88 · Founder & CEO
Signed ·
|