| name | ecommerce-conversion-optimisation |
| description | Use when the main deliverable concerns e-commerce and WhatsApp conversion diagnosis, buyer friction, and test prioritisation; use ecommerce-brand-differentiation when that neighbouring workflow owns the primary decision. |
| metadata | {"portable":true,"compatible_with":["claude-code","codex"]} |
Ecommerce Conversion Optimisation
Use When
- Use this skill for e-commerce and WhatsApp conversion diagnosis, buyer friction, and test prioritisation.
- Use it when the requested deliverable needs the domain decisions and acceptance checks below.
Do Not Use When
- Use
ecommerce-brand-differentiation when that neighbouring workflow owns the main decision or deliverable.
- Do not proceed when required evidence, approval, or safety review is absent; return the missing-input path instead.
Required Inputs
| Artefact | Source/provider | Required? | If absent |
|---|
| Objective, audience, market, and intended decision | Client or approved brief | yes | Ask for it or state a narrow working assumption |
| Existing channel, content, commercial, or performance evidence relevant to e-commerce and WhatsApp conversion diagnosis, buyer friction, and test prioritisation | Client systems, supplied files, or verified research | conditional | Mark the check unassessed and avoid performance claims |
| Approval, policy, budget, access, or risk constraints | Accountable client owner | conditional | Stop before publishing, spending, collecting data, or making regulated claims |
Workflow
- Confirm the decision, consumer, market, and evidence boundary; distinguish the request from
ecommerce-brand-differentiation.
- Inspect supplied artefacts and record missing or unverified inputs before drafting.
- Apply the domain framework in this skill and use the decision rule below at each branch.
- Stop for approval before publishing, spending, contacting people, changing live systems, or making regulated claims.
- Review the deliverable against the quality and anti-slop gates; if a check fails, correct it and rerun the affected check.
- Hand off the artefacts, assumptions, evidence, and unresolved risks to the named consumer.
Outputs
| Artefact | Consumer | Observable acceptance condition |
|---|
| E-commerce and whatsapp conversion diagnosis, buyer friction, and test prioritisation deliverable | Client decision-maker or delivery team | Names the chosen route, owners, sequence, assumptions, and measurable acceptance checks |
| Decision and risk record | Reviewer or implementer | Links each recommendation to supplied evidence or labels it as an assumption |
Evidence Produced
| Evidence | Format | Acceptance condition |
|---|
| Input and assumption register | Table or annotated brief | Missing and unverified items are visible, not treated as passed |
| Release check | Completed quality checklist | All blocking findings are fixed or the deliverable is explicitly withheld |
Capability and Permission Boundaries
Read and search are the minimum capabilities. Analysis and planning remain read-only. Edit only files placed in scope; publishing, outreach, spend, personal-data processing, production changes, and certification claims require explicit authority and evidence of success.
Degraded Mode
If files, tools, network, current evidence, rendering, or authorised access are unavailable, return the narrowest useful qualified deliverable. Mark each unavailable check not assessed; never convert it into a pass or invent market facts.
Decision Rules
| Choice condition | Action | Failure or risk avoided |
|---|
| Traffic exists but purchase or enquiry completion is weak | Diagnose the highest-evidence friction and specify one measurable test at a time | Changing brand strategy when the actual loss is in the transaction path |
| Evidence is contradictory or materially incomplete | Pause the affected recommendation and request the accountable source | Confident advice built on an unresolved premise |
| Authority is limited to analysis or planning | Deliver a read-only plan and approval checklist | Unauthorised publication, spend, outreach, or data use |
Quality Standards
- Keep Uganda/East Africa, British English, EAT, UGX, and WhatsApp-first assumptions explicit where they apply.
- Tie recommendations to observed evidence, a named assumption, or a verification action.
- Give the next operator enough detail to execute without guessing ownership, sequence, or acceptance.
- Apply
ai-marketing/anti-ai-slop during drafting and block release on an F from ai-marketing/ai-slop-audit.
Anti-Patterns
- Inventing a client metric, audience fact, price, partner, or platform rule. Fix: verify it or label the decision provisional.
- Treating a missing tool, source, render, or approval as a passed check. Fix: mark it
not assessed and narrow the output.
- Producing channel tactics before defining the decision and consumer. Fix: state the required outcome and handoff first.
- Copying a global template without adapting Uganda/East Africa access, language, payment, or trust conditions. Fix: record which local assumptions apply.
- Recommending publication, outreach, spend, data collection, or a regulated claim without authority. Fix: stop at an approval-ready draft.
- Reporting activity as success without an acceptance condition. Fix: name the observable result and evidence source.
References
Required Input
Ask the client for the following before generating any deliverable:
- Client business name and industry — e.g., "Kampala Fresh, organic food delivery"
- Country/city — default is Uganda/East Africa if not specified
- Current traffic sources — organic social, paid ads, WhatsApp broadcasts, referrals, walk-in, or a combination
- Approximate monthly enquiries and conversion rate — how many enquiries per month; how many convert to paid orders?
- Average order value (AOV) — in UGX or USD
- Primary drop-off point — where are most customers lost? (product page, WhatsApp enquiry stage, payment step, post-quote abandonment?)
- Existing tracking in place — Google Analytics, Meta Pixel, UTM parameters, manual tracking in a spreadsheet, or none?
- Primary goal — e.g., increase conversion rate from 5% to 15%, reduce cart abandonment, increase AOV, improve WhatsApp close rate
Section 1 — The Marketing Optimization System (MOS)
The Marketing Optimization System (Harris, 2016) is a three-module framework for diagnosing and fixing conversion problems:
- Customer Mindset — understand how visitors think, decide, and buy
- Gathering Intelligence — collect qualitative and quantitative data on actual behaviour
- Optimisation Process — use the 5-step process to test and scale improvements
Do not skip to solutions. Diagnosis comes before treatment.
Section 2 — Customer Mindset
4 Buyer Modalities
Every customer approaches a purchase through one of four decision-making styles (Harris, 2016). Effective product pages and WhatsApp scripts serve all four.
| Modality | Speed | Driver | What They Need |
|---|
| Competitive | Fast | Logic | Results, proof, performance claims, specifications |
| Spontaneous | Fast | Emotion | Urgency, excitement, FOMO, striking visuals |
| Methodical | Slow | Logic | Detailed FAQs, ingredient/component lists, comparison tables |
| Humanistic | Slow | Emotion | Founder story, community impact, testimonials, relationship |
EA audience default: Most Ugandan buyers trend Humanistic and Spontaneous. Open with trust signals and emotional resonance; provide the logical detail needed to close higher-value purchases (above UGX 200,000).
Application to content and copy:
- Product captions should open with an emotional hook (Spontaneous/Humanistic) and close with specific proof or specifications (Competitive/Methodical)
- WhatsApp scripts should include a brief story element or social proof before presenting the price
- Product pages and catalogue entries should have both a short emotional headline and a bullet list of specifications
Buyer Legends
A Buyer Legend (Harris, 2016) is the narrative gap between what your brand intends to communicate and what customers actually experience. To find the gap:
- Write out the ideal customer journey from first seeing a post to sending payment confirmation
- Walk the same journey as a new customer would — from a fresh social media account or via a competitor's ad funnel
- Document every point of confusion, missing information, or friction
- Each friction point is a conversion optimisation opportunity
Ideal Click M.A.P. (Marketing Along a Path)
Map the full funnel from first touchpoint to post-purchase:
| Stage | Touchpoint | Conversion Event |
|---|
| Awareness | Social post, ad, referral | View product |
| Interest | Product page, Stories, catalogue | Send WhatsApp enquiry |
| Consideration | WhatsApp conversation, quote | Request payment details |
| Decision | Payment instructions sent | Payment confirmed |
| Retention | Post-purchase message | Second order |
For each stage, identify the drop-off rate and the most common reason for drop-off.
Section 3 — Gathering Intelligence
Qualitative Research Tools
Qualitative research reveals why customers behave as they do. Use these methods:
| Tool | What It Reveals | Free/Paid |
|---|
| WhatsApp exit surveys | Why customers enquired but didn't buy | Free |
| Customer interviews (3–5 per month) | Emotional motivations, objections, decision process | Free |
| HotJar (heatmaps, session recordings) | Where on a product page attention drops | Freemium |
| SurveyMonkey or Google Forms | Structured buyer feedback | Free |
| Competitor funnel hacking | Become a customer of 2–3 competitors; document their full journey | Time cost only |
Qualitative research generates hypotheses. Quantitative data validates them.
Quantitative Research Tools
| Tool | What It Reveals |
|---|
| Google Analytics (GA4) | Traffic sources, bounce rate, session duration, conversion funnel |
| Meta Pixel | Facebook/Instagram ad performance, retargeting audience data |
| WhatsApp Business analytics | Message open rates, response rate, catalogue views |
| Manual spreadsheet tracking | Order-level conversion rate, enquiry source, AOV by product |
Minimum viable tracking for EA social commerce: A manual WhatsApp tracking sheet is sufficient for businesses under 50 orders per month. Log: enquiry source (which post or ad?), product enquired about, converted or not, reason if not converted. Review monthly.
Section 4 — The 5-Step Optimisation Process
Apply this process for every conversion problem identified (Harris, 2016):
Step 1 — Discovery: Combine qualitative and quantitative data to identify the highest-impact friction point. Prioritise by volume × severity.
Step 2 — Hypothesis: Form a specific, testable statement. "Changing the product caption from price-first to story-first will increase WhatsApp enquiry rate by 20% within 4 weeks."
Step 3 — Execution: Implement the change in one variable at a time. For social commerce: test one caption format, one CTA style, one Story layout, or one WhatsApp script element. Run the test for a minimum of 2 weeks or 200 impressions, whichever comes first.
Step 4 — Review: Compare the results against the baseline. Did the metric improve, decline, or stay flat? Require statistical significance (a minimum 95% confidence that the result is not random) before declaring a winner. For small-volume businesses, look for at least 20% relative improvement sustained over 2 weeks.
Step 5 — Scale: Apply the winning variant across all similar content or touchpoints. Document the winning pattern so it becomes the new default. Then return to Step 1 with the next friction point.
Section 5 — Key Conversion Metrics and Benchmarks
Core KPIs
| Metric | Definition | EA Social Commerce Benchmark |
|---|
| Enquiry conversion rate | Orders ÷ WhatsApp enquiries | Target: 30–50% |
| Post-reach-to-enquiry rate | Enquiries ÷ post reach | Target: 1–3% for product posts |
| Abandoned enquiry rate | Enquiries with no follow-up response | Target: below 15% |
| Average order value (AOV) | Total revenue ÷ number of orders | Aim to lift by 15–20% via upsells |
| Repeat purchase rate | Repeat customers ÷ total customers | Target: 40%+ at 6 months |
| Bounce rate (web/link pages) | Single-page exits ÷ total visits | Target: below 55% |
Abandoned Cart and Enquiry Recovery
An enquiry that did not convert is not lost — it is warm traffic (Larsson, 2016). Apply a structured recovery process:
- Within 30 minutes: send a WhatsApp follow-up — "Hi [name], just checking if you had any questions about [product]?"
- Within 24 hours: resend the product image with a specific testimonial or trust signal
- At 48 hours: offer an incentive — free delivery, a bonus item, or a 10% discount code
- After 7 days: move to the inactive broadcast list for monthly re-engagement
Target: recover 10–20% of abandoned enquiries through systematic follow-up.
Section 6 — Conversion Tricks and Revenue Boosters
Apply these proven tactics (Larsson, 2016) to increase conversion rate and average order value:
Retargeting
- Install the Meta Pixel on any website or link-in-bio page to capture all visitors
- Visitors are 70% more likely to convert after seeing a retargeting ad
- Use retargeting to show a product-specific ad to visitors who viewed that product page but did not enquire
Landing Pages
- A dedicated product landing page (even a single WhatsApp-linked link-in-bio page) converts 5–10% better than sending buyers directly to a profile
- Required elements: emotional headline, product images (minimum 3), key features and benefits, at least one testimonial, price, and a single clear CTA
Mobile Checkout Optimisation
- Simplify the WhatsApp order form to the minimum required fields (name, item, delivery address, payment method)
- Offer a payment link (Pesapal) for customers who prefer to pay without manually completing a Mobile Money transfer
- Guest checkout principle: never require registration, accounts, or multiple steps before a customer can pay
Free Shipping Threshold
- Set free delivery at 20% above current AOV to incentivise larger baskets without significantly increasing average fulfilment cost
- Announce the threshold in every product post: "Free delivery on orders above UGX [amount]"
Loss Leaders and Tripwires
- A low-cost or zero-margin entry product (loss leader) acquires a first-time customer; upsell the margin-positive product in the follow-up
- A tripwire is a deeply discounted, genuinely valuable first offer that converts cold traffic into buyers at scale
Flash Sales
- 24-hour flash sales with countdown timers in Stories generate urgency and compress the decision cycle
- Run flash sales on slow-moving inventory first; apply the discount to the least popular 20% of SKUs, not the bestsellers
- Announce 24 hours in advance, then remind at 6 hours, 2 hours, and 30 minutes remaining
Ride-Alongs
- Include a printed insert, a small bonus item, or a next-order discount code in every physical delivery
- Cost: minimal; effect: 30–40% take-rate on a repeat purchase incentive (Larsson, 2016)
Section 7 — Ecommerce KPI Dashboard
Define and track these KPIs monthly. Report to the client using simple visualisations (bar chart for trends, table for current-period values):
Revenue KPIs:
- Total revenue, revenue by product, revenue by channel
- Gross margin % by product (target: 60%+)
- Operating margin (EBITDA) — track daily for any business doing 20+ orders per day
Customer KPIs:
- New customers vs. repeat customers
- Customer acquisition cost (ad spend ÷ new customers from ads)
- Customer lifetime value (LTV) — average revenue per customer over 12 months
Conversion KPIs:
- Enquiry conversion rate (orders ÷ enquiries)
- Abandoned enquiry rate
- AOV and trend over time
Traffic KPIs:
- Enquiries by source (which post, ad, or channel generated the enquiry?)
- Post reach-to-enquiry rate by content type
Apply the RASTA standard to all client reporting (Phillips, 2015): Relevant, Accurate, Simple, Timely, Annotated.
Quality Criteria
Output from this skill meets the standard if it:
- Diagnoses the specific conversion bottleneck before recommending solutions
- Segments visitors into the 4 buyer modalities and adapts copy recommendations accordingly
- Maps the full customer journey from first post to payment confirmation, identifying each drop-off point
- Includes at least 3 specific conversion optimisation hypotheses with measurable success criteria
- Provides a 5-step testing plan (Discovery → Hypothesis → Execution → Review → Scale) for the top priority friction point
- Sets a KPI dashboard with at least 5 tracked metrics and baseline benchmarks for the EA market
- Includes a structured WhatsApp abandoned enquiry recovery sequence
- Names specific tools appropriate for the client's scale (manual spreadsheet for small volumes; HotJar, Meta Pixel, GA4 for larger operations)
References
- Harris, A. (2016) Small Business Big Money Online. Marketing Optimization System, 4 Buyer Modalities, 5-Step Process
- Larsson, T. (2016) Ecommerce Evolved. Conversion tricks, traffic temperature, retargeting, flash sales, ride-alongs
- Phillips, J. (2015) Ecommerce Analytics. KPI frameworks, dashboarding, RASTA reporting
social-commerce-strategy/SKILL.md — EA-specific social commerce operations and platform setupecommerce-brand-differentiation/SKILL.md — Brand positioning and intangibles strategy
