Website audit & build proposal

Mwanangu Baby Wear

A rebuild of the shop front so the marketing has somewhere to land — keeping the parts of the current build that already work.

Prepared for
Mwanangu Baby Wear, Harare
By
MSN Technologies
Date
9 September 2026
Site reviewed
mwanangu.shop

01The short version

The engine is fine. The shop window is invisible.

Whoever built this did the hard parts properly — Paynow payments with EcoCash and cards, product variants, stock tracking, 63 delivery zones already priced, a working admin. That is real work and it should not be thrown away.

What is broken is everything a marketer needs. The site renders entirely in the browser, so Google, Facebook and WhatsApp all receive a blank page. Share a product link on WhatsApp today and it appears as a grey box with no picture and no price. For a Zimbabwean baby store, where almost every sale starts as a WhatsApp forward, that single fault undoes the rest.

02What I found

Checked live on 9 September 2026 against the public site and its API. Every item below is verifiable, not an impression.

Critical
Nothing is rendered on the server

Every page — home, product, category — returns the same empty <div id="app"></div>. Search engines and social crawlers do not run JavaScript reliably, so they see an empty document with no products, prices or descriptions.

Critical
No link previews anywhere

There is no meta description and no Open Graph tag on any page. A link pasted into WhatsApp, Facebook or Instagram shows no image, no product name and no price. This is the biggest single obstacle to the marketing you are about to run.

Critical
Every page has the same title

All 61 products and both categories return Mwanangu Baby Wear. Google has nothing to tell one page from another, so it will rank at most one of them.

Critical
robots.txt and sitemap.xml do not exist

Both addresses return the app's HTML page instead of a 404. Google asks for a sitemap and receives a web page — so no product URL is ever submitted for indexing.

Critical
No analytics and no Meta Pixel

Nothing is installed. There is no way to see what people look at, where they abandon the cart, or which advert produced a sale — and no way to retarget anyone who visited. Advertising spend would be flying blind.

High
39 of the 61 products have no photograph

Only 22 products carry an image. Nobody buys baby clothes they cannot see, so almost two thirds of the catalogue is effectively unsellable.

High
You cannot shop by age or size

The only navigation is Boys and Girls. A parent who needs a 2-year-old's outfit has to open products one at a time. Age is the first question every parent asks and the site cannot answer it.

High
The homepage promises delivery that is switched off

The hero reads “delivered to your door” and the footer strip says “Delivery across Zimbabwe”, but the store setting is delivery_enabled: false — the site is collection-only at 49 Cork Road. Meanwhile 63 delivery zones are already configured at $4 to $18. Either turn it on or change the words; right now the site contradicts itself.

High
“Pay with EcoCash & Cards” while ZiG is disabled

ZiG payments are off (zig_payments_enabled: false) though a rate of 36 is stored. Prices show in USD only.

Medium
No reviews and no email capture

Nothing collects proof or an audience. There is no way to build a list you own, so every campaign has to be paid for again.

Medium
Brands are barely used

Two brands across four products. Woolies and Ackermans are trusted names in this market and are worth merchandising properly.

03What already works — and stays

Do not rebuild these

  • Paynow payments — EcoCash, OneMoney, ZIPIT, Visa and Mastercard, correctly wired with an initiate-and-poll flow.
  • Product model — variants, stock levels, brands, categories, product types, dual USD/ZiG pricing.
  • Stock movements — a proper audit trail, not just a counter.
  • 63 delivery zones with per-suburb fees from $4 to $18, already captured.
  • Admin dashboard — products, orders, stock, settings, hero slides.
  • The photography. The 22 product shots that exist are genuinely good — consistent light, clean flat-lays. The problem is that there are not enough of them.

04Scope and effort

Estimated in working days for the recommended option.

WorkDays
Discovery, API mapping, content and size taxonomy2.0
Design system and homepage (per the demo)4.0
Shop listing: filters by age, category, brand, price3.0
Product detail: gallery, variants, size guide, related2.5
Cart, checkout, Paynow handoff, delivery zones3.0
Account, order history, order tracking2.0
SEO layer: SSR, metadata, OG images, JSON-LD, sitemap, robots2.5
Analytics, Meta Pixel, product feeds for Meta and Google1.5
Email capture and reviews1.0
QA, mobile, performance, launch and handover2.5
Total24.0 days

05Three ways to do it

Rescue

8 days
R10 000
  • Server rendering added to the existing site
  • Metadata and link previews on every page
  • Sitemap, robots, page titles
  • GA4 and Meta Pixel
  • Delivery and currency copy corrected
  • Keeps the current design
Recommended

Rebuild the shop front

24 days
R14 000
  • Everything in Rescue
  • New storefront in Next.js, per the demo
  • Shop by age and size
  • Product feeds for Instagram and Facebook Shopping
  • WhatsApp ordering and cart sharing
  • Email capture and reviews
  • Keeps her database, admin and Paynow

Rebuild everything

36 days
R18 000
  • Everything above
  • New backend and database you own outright
  • All 61 products migrated
  • New admin
  • No dependency on the previous developer
  • Choose this only if handover is refused

Ongoing

Care plan — hosting, backups, platform updates, up to 2 hours of support a monthR1 250 / month

Support covers things like a price correction, a swapped photo, or new stock — not new features. Unused hours don’t roll over; anything beyond 2 hours in a month is billed at an hourly rate agreed upfront.

06Timeline

WeekWhat happens
Week 1Discovery, access handover, size and age taxonomy agreed, design system signed off
Week 2Homepage and shop listing built and reviewed
Week 3Product detail, cart, checkout, Paynow verified end to end
Week 4Account and tracking, SEO layer, analytics and feeds
Week 5Content load, photography gaps filled, QA, launch

Assumes photography and product copy arrive by the end of week 2. That is normally the thing that slips.

07What I need before I can firm this up

  1. Does she control the server, domain and source code? This is the one that changes the price. If the previous developer will not hand over the repository and server access, the only honest option is the full rebuild. Everything else assumes cooperation.
  2. Is she doing delivery or not? 63 zones are priced and ready but the switch is off while the homepage promises delivery. Which is true?
  3. Does she have age and size data per product? Shop-by-age is the biggest single conversion win available, and it needs each product tagged. If it is not recorded anywhere, that is a data-entry job for 61 products.
  4. Can we photograph the 39 products with no image? Are they still in stock, and where are they physically?
  5. How many orders a month at the moment, and what sells? This sets whether the marketing budget goes to acquisition or to repeat purchase.
  6. What are the Instagram and Facebook accounts, and the follower counts? Also whether a WhatsApp Business catalogue already exists.
  7. What is her budget and her deadline? If there is a season she is aiming at, say so and the scope gets cut to fit rather than the quality.

The demo

A working homepage concept, built with her real catalogue, her real photography and her real prices — not a template with stock images. Open it on a phone; that is where her customers are.

View the homepage demo →

Mwanangu Baby Wear — audit and proposal MSN Technologies · 9 September 2026