AI lead scoring with reasons your team can read — in Arabic and English.
Every SOOMA AI lead score comes with five plain-language reasons translated into both Arabic and English. Reps know why the score is what it is. Managers coach with evidence. Nobody trusts a black box — so we built one that talks.
The problem with most AI scoring
The CRMs you've evaluated (Salesforce Einstein, HubSpot, Zoho ZIA, Freshworks Freddy) all score leads with black-box models. They give you a number; they don't tell you why. In English-only product UIs, that's already a coaching problem. In a UAE field-sales context — where the score comes back in English and the rep speaks Arabic — it's a complete dead end.
Reps don't trust scores they can't explain to themselves. Managers can't coach on signals they can't reason about. Pipeline reviews turn into "the AI said so" instead of "here's the factor that matters."
How SOOMA's scoring works
Every lead gets a 0–100 score updated in milliseconds. Alongside the score we surface the top five contributing factors using a SHAP-style attribution model — the same explainability technique used in production ML at Microsoft, Uber, and Airbnb — translated into plain Arabic and plain English.
Typical reasons your team will see:
- +15 Reached the demo stage (وصل لمرحلة العرض)
- +12 Deal value above the org average (قيمة الصفقة أعلى من المتوسط)
- +8 Strong digital presence (حضور رقمي قوي)
- +6 Active in the last 7 days (نشط خلال 7 أيام)
- -2 Industry below average win-rate (قطاع أقل من معدل الفوز)
What signals SOOMA actually scores
- Deal value vs your org's historical average. A AED 12K lead in a AED 3K-average org gets a different weight than the same lead in an AED 50K-average org.
- Pipeline stage progression. Time-to-stage and stage velocity, not just current stage.
- Engagement recency. Last activity timestamp normalized to your sales cycle length.
- Digital footprint. Enrichment data: website, social, employee count, industry.
- Industry conversion. Your org's win-rate by industry, not a global benchmark.
- Rep-level conversion. Some reps close certain profiles better — the score knows.
Why Arabic SHAP matters
A score with English-only reasons is invisible to the rep who needs to act on it. We translate every factor into the rep's preferred language at render time — not pre-baked translations from a fixed table, but context-aware Arabic that matches the actual factor magnitude and direction.
The result: pipeline reviews where managers and reps argue about the right things ("the deal value is right but the industry conversion is weak") instead of arguing about the score itself.
Trust and compliance
The scoring model runs in our UAE-hosted environment. No lead data leaves the region. PDPL-aligned audit logs capture every score change and the inputs that drove it. See our Trust page for the full control list and PDPL-compliant CRM for the legal context.
How to evaluate it
We don't offer a generic free trial because lead scoring with no historical deals is just guessing. Instead, book a 30-minute demo and we'll score 200 of your real leads (anonymized, of course) before the call. You decide whether the reasons match your gut.