Prepared with love, not with invoices
Your dream. Our homework.
You brought us KASSIE. We wish we had the hands to build it with you. We don’t right now. So here is the next best thing: what we would want to know before writing a single line of code, in plain language, with real numbers, yours to hand to whoever builds it.
One minute, tops
The short version
One checkout across many shops is real for independent boutiques that run on Shopify, the system most small online shops use. The big names in your demo pages (Poetry, Country Road, Old Khaki, Totalsports, Yuppiechef) each run their own system. Every one of those needs a signed deal, then a custom-built connection, before it can appear in a Kassie. Your job is signing shops. The engineer’s job is plugging them in.
We looked around
You've spotted something real
We read every public page on mykassie.com (home, browse, rewards, the Winter Wardrobe demo, the dashboard) and then went looking at how the rest of the world has tried this.
Even Google can't pay once yet
Google launched a “universal cart” this year. Shoppers can collect items from many stores. When they pay, each store still takes its own payment. Google is not the seller. That last step, one payment across different businesses, is still on their wish list.
The overseas giants dodge it
The big creator-shopping apps abroad let people curate, then send shoppers out to each store’s own till. Nobody has cracked one-cart curation as a product. That is your opening, and your warning about how hard it is.
You can clearly design
The KASSIE site is warm, clear and human. That matters more than it sounds: it means you can be the product and design lead yourself, which roughly halves the build cost.
What the live site actually is, today
A Johannesburg waitlist for a three-sided marketplace. Shoppers browse curated collections. Creators earn 4–14% on referred sales. Retailers list a catalogue and pay commission only on tracked sales. The headline promise: one outfit, every retailer, one checkout, one shipping fee, with each retailer still posting its own parcel.
Your own footer is honest: private beta, no retailers signed, nothing charges a real card. Launch is spoken of as 2027. The only named retail partner is Hannah Collective, which already runs on Shopify. Everything else (the Winter Wardrobe, the dashboard numbers, the R4.2 million “total value shopped”) is labelled as illustration or as a goal. That is fine. It just means the hard work is still ahead, and most of it is not software.
The honest bit
What we love, and what's genuinely hard
What we love
- A real gap: South African creators drive sales every day and earn almost nothing from them. Your fix is fair to all three sides.
- Retailers pay only when they actually sell something, an easy yes for a small boutique.
- The monthly shop fee (R500–R1 500) means real income from the very first shops, before sales volume exists.
- Hannah Collective already runs on Shopify, connectable from day one, once they say yes.
- The brand is already good. Warm, clear, human. Do not let an agency redo it.
What’s genuinely hard
- No ready-made “shared cart” exists anywhere. KASSIE must build and own its own checkout. Doable, but it is the core of the build.
- Whoever takes the money owns the refunds, returns and complaints for the whole order, across every shop in it. A lawyer needs to shape this in month one, not month twelve.
- “One shipping fee” with four parcels means someone pays the other three couriers. That is a business decision, not a software feature.
- 2% shopper cashback for life plus creator commission cannot both come out of the same retailer commission. The maths below shows why.
- Each big retail group is 6–12 months of talks before any code helps. Their logos sell the dream, but they cannot be the launch plan.
The one big catch
The shops in your demos mostly can't plug in yet
Winter Wardrobe lists Poetry, Country Road, Old Khaki and Aldo. The dashboard also names Totalsports, Sheet Street, Yuppiechef, Zara and iStore. Lovely collections, and none of those groups run on Shopify. Each one needs a signed deal, then a custom connection that your engineer builds and keeps alive. This is the single most important thing to plan around.
Independent boutiques: Hannah Collective, indie beauty and home brands
Their systems: Mostly Shopify
When they can join: Day one. They install KASSIE's add-on and their catalogue flows in
Small shops on other common systems
Their systems: Other popular shop software
When they can join: Later. One extra connection type, once the Shopify version works
Poetry, Old Khaki, Country Road, Totalsports, Sheet Street, Woolworths
Their systems: Big-group platforms, each different
When they can join: Only after a signed deal, then a custom connection per group
Yuppiechef, iStore and other large independents
Their systems: Their own custom systems
When they can join: By agreement: a deal plus a custom connection, one at a time
Wait, what exactly is an “API”?
An API is the plug on the side of a shop’s software that lets other software connect to it: to read the catalogue, check stock, and place orders. Shopify gives every shop the same standard plug, which is why thousands of boutiques can be connected with one piece of work. The big retail groups each have their own differently shaped plug (or none they will share), which is why each one is a separate project, after they sign.
Under the bonnet
You sign the shops. They connect them.
If this were “just Shopify”, a good engineer could pull it off as a neat add-on. The shops you want to attract do not all live there. So the shape of the work is: you bring people in, they build the pipe. Start with Shopify boutiques. Add the giants only when a named person at the group has said yes in writing.
- 1
You sign a boutique
A one-page yes: they install KASSIE's small add-on, pay around 8% on tracked sales, keep their own packing and posting. This is the work only you can do.
- 2
The engineer plugs them in
Their whole catalogue (products, prices, stock) flows into KASSIE and stays in sync. One Shopify connection covers every boutique on Shopify. A Poetry-sized group is a new project, after they sign.
- 3
Creators curate
Creators search everything that's connected, build a Kassie, and share one link. This part is yours already. It's the heart of the brand.
- 4
The shopper pays once
On KASSIE's own checkout, using South African payment services like Paystack or PayFast. One card payment, exactly as your homepage promises.
- 5
KASSIE splits the order
Behind the scenes, each shop in the cart instantly receives its part as a normal order in its own system, marked as paid. They pack and post as usual. Parcels arrive separately, which your site already says, honestly.
- 6
The money settles
Each shop is paid its share, the creator is paid their commission, and KASSIE keeps its fee. A clean weekly rhythm works fine at the start.
Why can’t the shops just share one cart?
Every online store’s checkout is its own till. Shopify, and everyone else, simply does not let separate businesses ring up one combined payment. Google is trying to build exactly this right now and has not finished either.
The workaround is the whole product: KASSIE runs its own till, takes the one payment, then quietly places a separate paid order with each shop. To the shopper it feels like one checkout. To the shops it looks like normal orders arriving.
What does “taking the payment” really mean?
If the shopper’s card statement says KASSIE, then in the eyes of South African consumer law, KASSIE is the seller. Refunds, returns, complaints and chargebacks for the whole order land on KASSIE first, even when the late parcel is one boutique’s fault.
This is manageable with clear retailer contracts (a lawyer can structure KASSIE as an agent selling on each shop’s behalf), but it must be set up in month one. It is the least glamorous and most important piece of the whole project.
And the “one shipping fee” promise?
Four shops in a cart means four couriers get paid. If the shopper pays one flat fee, the difference comes from somewhere. Two clean options: retailers cover their own shipping on KASSIE orders (a condition of joining), or KASSIE charges a flat fee (say R79) and absorbs the difference as a marketing cost it tracks honestly. Pick one before launch; do not discover it in the bank statement.
Is there a cheaper first step if one checkout can wait?
Yes. A tracked-link version (creators curate, shoppers click through to each shop’s own till) can be up in a couple of months for a few hundred thousand rand. It works with anyone who already has an affiliate programme, including the big names. It is useful. It is not the product on your homepage.
We would only take that path if you decide one checkout can wait. If the homepage promise is the point, build the boutique version of the real thing.
The big rethink
You are the product and design team
A typical quote for this assumes three salaries: two engineers plus a product-and-design lead. The KASSIE site shows you are the product-and-design lead. Strike that salary. Then pick a path. The numbers below move with your choice.
The punchline: a hired three-person team (two engineers plus a product-and-design lead) comes out at roughly R2 500 000 for the first paying version, and about R6 750 000 for the full dream. With you doing product and design, one great engineer lands the first version at about R1 200 000: a little slower, because one person is doing all of it. Your eye is worth roughly a million rand. Spend it.
Costs update as you pick a path
Mostly your time, plus a good lawyer. Ten boutiques on a one-page agreement. Do not hire until those yeses are real.
7–8 months of one exceptional engineer. Catalogue sync, creator boards, your own checkout, order splitting, simple weekly payouts. No points, no draws.
Refunds, live stock checks, retailer dashboards, wallet withdrawals, sizes done properly. Another four to five months.
Points, referrals, monthly draws, gift lists, and the next shop system, once the first fifty real orders have cleared.
Each big retailer, when the time comes
R0.8–3.5 million per group, plus 6–12 months of talks, and the talks come first. Budget nothing here until a named person at the retailer owns the project in writing. Then your engineer builds that one connection.
Not in these numbers
Legal advice (start month one), hosting (small: R8–25k a month), card fees of about 3.5% on every sale, whatever the one-shipping-fee promise costs, and keeping shop connections up to date. Plan 15–25% of the build cost each year for upkeep.
From a cost perspective
Can the shop fees pay the engineer?
Retailers pay KASSIE R500, R1 000 or R1 500 a month, plus commission only when something sells. That monthly fee is the only dependable income before sales volume exists. Play with how many shops you sign.
- 10 shops × R1 000
- R10 000
- This hire costs
- R150 000 / month
You would need about 150 shops at R1 000 a month before fees alone cover this hire. Ten founding boutiques at R1 000 is R10 000. Lovely, and nowhere near R150 000. Plan runway for the build, then let shops and sales grow into the salary.
Play with it
The money on one sale
This is the sum to do before writing any code, so we made it a toy. Drag the slider, switch the creator tier, and watch where the money goes. The short version: creators are paid generously, and KASSIE keeps a sliver. That sliver does not cover card fees, let alone cashback.
- The retailer pays KASSIE (8–10% (we use 9%) commission)
- R90
- The creator is paid out (commission minus KASSIE's 3% cut)
- R87
- KASSIE keeps as its fee
- R3
- Card and payment processing (about 3.5% of the sale)
- − R35
- Cashback paid to the shopper (2%)
- − R20
See the problem? On these numbers KASSIE loses money on every sale. The gap has to be closed somewhere: the retailer’s monthly fee (R500–R1 500) helps, card fees can be pushed to the retailer’s side in the contract, and the cashback needs a clear funder. None of this is fatal. It just needs deciding before anything is promised “for life”.
The most expensive sentence on your site is “founding shoppers earn 2% cashback for life”. Creator commission is funded by the retailer. Great. Cashback stacked on top comes out of KASSIE’s thin slice, and it does not fit. Decide who funds cashback (a retailer top-up, a smaller rate, or points instead of cash) before the founding cohort grows.
The plan
The order to do things in
- 1
Sign ten boutiques first
A one-page agreement: they install the add-on, pay around 8% on tracked sales, keep their own fulfilment. Do not approach the Poetry-sized names yet. You will get one shot at them, and it should be with a working product.
- 2
Decide the two money rules, in writing
Who funds shopper cashback, and how the one-shipping-fee promise is paid for. Both are business decisions that shape the build, and both are currently promised 'for life' on the site.
- 3
Hire, then build the lean version
Catalogue sync, creator boards, your own checkout, order splitting, simple payouts. Skip points, draws and tiers for now. They are lovely, and they can all wait.
- 4
First 50 real orders, then expand
Once real money has cleared 50 times, add the rewards layer, connect the next shop system, and start the big-retailer conversations from a position of proof. Your engineer adds each new pipe as you bring the people in.
Take this to the interview
How to spot the right engineer
The unicorn is hard to find. If someone has actually wired up several different store systems, has a good eye, and is happy to build from your designs: that is the person. If you cannot find them, take Path B. Do not settle for a generalist who has never touched a shop’s plug.
Three questions to ask
- “Show me something you’ve built that talks to several different store or payment systems.” You want war stories, not theory.
- “Have you taken real money with Paystack or PayFast?” South African payments have quirks. Experience here saves months.
- “A shopper pays for a three-shop cart and one item just sold out. What happens?” A good answer mentions re-checking stock at the moment of payment, and a plan for refunding just that item gracefully.
Green flags
- Has shipped a marketplace, booking or payments product end to end
- Asks about refunds and returns before you bring them up
- Happy to build from your designs, and it shows in their past work
- Talks about starting small and cutting scope
Red flags
- Promises a shared cart across Shopify stores “out of the box”. That does not exist
- Wants to start with points, draws and gamification
- Quotes a fixed price without asking which systems the retailers run on
- Answers every hard question with “there’s a plugin for that”
With love
The bottom line
KASSIE is a genuinely good idea wearing a slightly-too-big coat. Tailor it: boutiques first, one great hire (or one great pair), your own design eye, and honest maths on cashback and shipping before anything is promised for life. The giants come later, by signature, not by software.
We cannot be the ones to build it, and we are a little sad about that. But whoever you hire, hand them this page and hold them to it. And when the first real Kassie ships four parcels from one checkout, we are buying one.
Go build it. ♥