Triggering Product Recommendations with Chatbots in Kenya.

Kenya is a mobile-first market where discovery, consideration, and purchase often happen on the phone. If you’re not making contextual product recommendations in the same channel customers use daily (SMS and chat), you’re leaving revenue on the table. With SMSLeopard and ChatSasa, you can trigger automated, hyper-relevant upsells and cross-sells at precisely the right moment—without adding headcount.

This guide shows how Kenyan businesses can design recommendation flows, choose high-signal triggers, localize content, and measure lift—end to end.

Why automated product recommendations matter (in Kenya)

• Mass reach: SMS works on feature phones and smartphones; WhatsApp/chat flows deepen engagement.

• Speed to value: Recommendations arrive right after intent (viewed product, cart event, delivery, refill date).

• Revenue lift: Timely upsells/cross-sells increase AOV (average order value) and repeat purchase rate.

• Localization: Multilingual messaging (English, Kiswahili, plus vernacular) improves trust and clarity across counties.

• Cost efficiency: Automations run 24/7, escalating to agents only when needed.

High-signal triggers that drive relevant recommendations

Use behavioral, transactional, and lifecycle triggers to time your suggestions:

1. Browse/interest signals

   • Clicked a category link in SMS/WhatsApp

   • Asked a bot about a product/spec

   • Viewed size/color variants in chat catalog

2. Cart & checkout signals

   • Added to cart but didn’t check out

   • Removed an item (offer a substitute)

   • Checked out (now suggest complementary items)

3. Transactional signals

   • Order shipped/delivered (accessories or refills)

   • Subscription renewal due (upgrade tier)

   • Warranty/return window (care plans, protection)

4. Lifecycle signals

   • 30/60/90-day inactivity

   • Seasonal demand (Back-to-School, Jamhuri Day, festive season)

   • Geography/branch proximity (recommend nearby offers)

5. Preference signals

   • Survey or quick poll responses (price vs speed vs quality)

   • Language preference (Kiswahili/vernacular)

   • Payment preference (M-Pesa vs card)

Pro tip: Start with 3–5 triggers that correlate strongly with purchase, then expand.

Building blocks in SMSLeopard & ChatSasa

• Segmentation: Group by county, language, product category, spend band, last action (clicked, carted, purchased).

• Dynamic templates: Insert {FirstName}, {Category}, {Branch}, {Voucher}, {PayLink}; maintain multilingual versions.

• ChatSasa bots: Guide users with buttons (e.g., “See Colors,” “Compare Plans,” “Add Accessory”).

• Link shortener + UTM: Track clicks and conversions per trigger, message, and segment.

• Automation & routing: Trigger flows instantly; escalate to human agents on intent keywords (“talk to agent”).

• Compliance controls: Built-in opt-out handling, consent capture, and DND filtering.

Blueprint: designing a recommendation flow

1. Trigger fires (e.g., cart created, delivery complete).

2. Eligibility check (opt-in, frequency cap, language, inventory).

3. Message 1 (primary rec) with 1–2 buttons: “View details”, “Add to cart”.

4. Wait window (e.g., 2–4 hours). If no action:

5. Message 2 (alternative rec) or small incentive (time-boxed).

6. Exit & tag (Converted / Deferred / Not interested).

7. Attribution (UTM click → page/product; measure CTR, CVR, AOV, revenue per send).

8. Agent takeover if user asks a complex question or types “agent”.

Industry playbooks (Kenya-ready)

1) Retail & E-commerce

Triggers: Viewed product, cart abandonment, delivery confirmation. Recommendations:

• Substitutes (same price/size/color)

• Complements (accessories, add-ons)

• Bundles (save with 2+ items)

Sample SMS (Kiswahili + English): “{FirstName}, umependa {Category}? Ongeza {Accessory} kwa 10% off leo. Reply 1 kuongeza / 2 kuona rangi.” “Hi {FirstName}—complete your look with {Accessory}. Tap to add and checkout now: {ShortLink}”

KPIs: AOV, cart recovery rate, time-to-purchase.

2) Pharmacies & Healthcare

Triggers: Rx refill due, clinic visit done, wellness campaign interest. Recommendations:

• Refill reminders + dosage aids

• Complementary wellness items (vitamins, devices)

• Care plans and tele-consults

WhatsApp flow: Buttons: “Refill now”, “View alternatives”, “Talk to pharmacist” If “Refill now” → M-Pesa pay link → order confirm SMS.

KPIs: Refill adherence, repeat purchase rate, clinical outcomes proxies (appointment kept).

3) Financial Services (SACCOs, MFIs, Banks)

Triggers: Loan eligibility change, salary date proximity, repayment completion, low balance. Recommendations:

• Higher-limit or faster-disbursement loans

• Savings products with bonus rate

• Insurance add-ons (device, travel, health)

SMS example: “{FirstName}, your good repayment history qualifies you for up to KES {Limit}. Need funds in 24 hrs? Reply 1 to view options.”

KPIs: Approval rate, uptake per segment, NPL impact.

4) Education (Schools, Universities, EdTech)

Triggers: Application started, fee due, exam timetable issued, course completed. Recommendations:

• Prep materials, revision bundles

• Transport/hostel options

• Next-level course upsells

ChatSasa flow: “Choose what you need: 1) Revision pack 2) Past papers 3) Coaching call” → Deliver links/resources instantly.

KPIs: Enrollment conversion, fee timeliness, course progression.

5) Logistics & Delivery

Triggers: Quote given, shipment delivered, SLA events. Recommendations:

• Insurance/priority handling

• Bulk discounts for recurring shipments

• Packaging supplies

SMS: “Parcel delivered to {Estate}. Save 12% on your next 3 shipments—activate a prepay bundle today: {ShortLink}”

KPIs: Repeat orders, average shipments per customer, bundle adoption.

Message patterns that convert (steal these)

• Cross-sell after purchase: “Asante, {FirstName}! Customers who bought {Item} also choose {Complement}. Add now & save 8% till 8pm: {ShortLink}”

• Substitute suggestion: “{Item} is low in stock at {Branch}. Prefer {AltItem}? It’s ready for pickup today. Reply 1 to reserve.”

• Value-led upsell: “Upgrade to {PlanPlus} for 2× speed + priority support. Free trial ends midnight. Tap to switch: {ShortLink}”

• Refill reminder (health): “{FirstName}, {Medication} refill due in 3 days. Reorder now for same-day pickup or delivery: {ShortLink}”

• Basket booster: “Add {Accessory} and unlock free delivery (saves KES {Amount}). Reply 1 to add, 2 to skip.”

Keep CTAs short and binary (1/2 buttons). Use Kiswahili/vernacular where it improves clarity.

Timing & frequency: get the psychology right

• Moments of intent: Immediately after browse/cart, or right after delivery.

• Pay cycles: Payday and early evening boosts discretionary purchases.

• Quiet hours: Respect rest times; schedule for morning/lunch/evening.

• Frequency caps: Limit recs to 1–2 per trigger per week; suppress after “Not interested”.

Data & content governance

• Source-of-truth: Sync product feeds (price, stock, images) to avoid recommending out-of-stock items.

• Eligibility rules: Exclude already purchased or incompatible items.

• Language accuracy: Pre-approve translated templates; test with native speakers.

• Compliance: Honor consent, opt-outs (“Reply STOP”), DND lists, and data minimization under Kenya’s Data Protection Act.

Measurement: prove lift, not just clicks

Track inside SMSLeopard (with UTM on links):

• CTR and conversion rate per trigger and template

• AOV and Revenue per 1,000 messages (RPM)

• Attachment rate (how often the rec is added)

• Time to purchase after message

• Incremental lift vs. a control group (no recommendation)

Run true A/Bs: same audience, different copy/offer; or same copy, different send times.

Quick start: 14-day rollout plan

Days 1–2: Pick 2 triggers (cart abandonment, post-delivery). Define segments and frequency caps. Days 3–4: Draft multilingual templates; wire buttons/keywords in ChatSasa. Days 5–6: Connect product feed; set eligibility and exclusions. Day 7: Enable UTM + short links; set up dashboards for CTR/CVR/AOV. Days 8–10: Launch to 20% of traffic; keep a clean control group. Days 11–12: Promote the top variant; pause underperformers. Days 13–14: Add a third trigger (refill/renewal). Document learnings.

Common pitfalls (and easy fixes)

• Too many options: Offer 1 primary recommendation + 1 alternative, not catalogs in SMS.

• Generic copy: Reference the exact item/category they showed interest in.

• No inventory check: Sync stock before sending offers.

• Over-messaging: Cap frequency; suppress after “Not interested.”

• No attribution: Always tag links; review revenue per send, not just CTR.

Why SMSLeopard + ChatSasa are your best bet

• Kenya-optimized delivery across Safaricom, Airtel, and Telkom.

• Unified channels (SMS, WhatsApp, web chat) in one dashboard.

• Behavior-based automation that triggers instantly on real events.

• Deep personalization by region, language, and shopping behavior.

• Analytics that matter with built-in short links and UTM-friendly reporting.

• Compliance-ready with consent capture, opt-out management, and DND filtering.

• Agent handoff from bot to human—seamless when conversations need a person.

Turn every interaction into the next purchase. Launch your first recommendation flow with SMSLeopard today. Visit www.smsleopard.com