Raava
Eight things we found in your own data, and what each one is worth. One finding at a time, and every number here comes out of a system you already own.
Website sales only. Marketplace sales are kept out, because they would hide what the website did.
A note on these figures These totals come from your Shopify order records, and they are gross of refunds. Your Shopify Analytics screen takes refunds back out, so it reads about 5% to 12% lower. Both are right. They are counting different things.
Website sales · twelve months at a time · marketplaces excluded
The scale starts at zero and nothing is cut off, so the drop and the recovery are the true size. The dotted line runs out of where you started, and the last point sits just above it. Two things happened over this window. You switched your marketplaces back on in late 2025, and this website line recovered alongside them. We have the month by month series for both.
School shoes is the most seasonal thing your shop sells. On your own website, January has been the biggest month of the year in each of the last five years.
School shoes · share of the category's year by month · your own website, 2021 to 2025
Each bar is that month's share of five years of school shoe sales through your own website.
A month with no season at all would sit at 8.3%. Taken year by year, January runs between
26.0% and 35.3% of the category's year across 2022 to 2025, and it has been the biggest month
of the year in each year from 2022 to 2026. We have not put a dollar on January alone: its
share is measured a year at a time and the sales figure covers twelve months, so the two
cannot be multiplied.
And right now those pages are not doing the selling. The one a parent reaches in two taps from
your own menu carries six words: "Shop our school collection!" Nothing about fitting, growing
feet, half sizes or widths. Five other school shoe pages carry a paragraph about women's boots.
Your biggest school shoe month is five months away, and the pages that sell school shoes do not describe them.
A$19,468School shoes altogether sold this on your own website in the last twelve months, up 27.6% on the year before.
Rewrite the school shoe pages so they describe school shoes, and finish before January rather than during it.
Check it in a minute Open your shop on a phone, tap the menu, tap School, and read the writing under the shoes. It is six words long. The five pages that describe women's boots, we will open with you.
Problem
The pages that carry school shoes do not describe school shoes, in the five months before your biggest month of the year.
Solution
Rewriting the school shoe pages is part of the same category-writing fix costed in the ninety-day proposal at twelve hours, which covers every wrong page on your shop at once, not this one alone. It can run as its own pass first, since January is the nearer date, or inside that twelve-hour pass with the rest.
Expected result
The school shoe pages describe school shoes, finished before January rather than during it. No confidence percentage: nobody has sampled a hundred shops like yours. What this rests on is five years of the same January pattern, on your own website.
WILDE is the largest of the nineteen brands your own shop sells, and it is growing. The page that holds every WILDE style exists, but your SEO setup is not indexing it properly, so almost no one outside the WILDE pages themselves can reach it.
Share of your own shop's sales · last twelve months · nineteen brands
Where your customers can reach it from
Google's own record for that page, in your Search Console export, is three clicks and 101 times shown, over the sixteen months to 28 July 2026. The page exists, but the SEO setup is not indexing it properly, so the way in is missing. Type Wilde into your own search box and Wilde shoes do come back. What no customer reaches is the one page holding all of them, and on a phone your Wilde logo lands on a page of Panda slippers.
Your biggest brand sits on a page nothing on your shop points to, and on a phone its logo on your homepage sends customers to a different brand's slippers.
A$60,782WILDE sales on your own website in the twelve months to 11 August 2026, up 58% on the twelve months before that.
Give that page a way in from the places customers already land, and treat WILDE as the front of the shop it already is.
Check it in a minute Open ishoes.com.au and try to reach the page that holds every WILDE style, using only your own menu. You will find Wilde pages. What you will not find is the one with all 159 of them on it.
Problem
Close to a third of your online sales sit on a page your shop has, but your SEO setup is not indexing properly, so the writing above cannot be the only fix.
Solution
Two separate tasks. Fix the SEO setup so the page is indexed and reachable. Rewriting the copy on the four Wilde pages is part of the same twelve-hour category fix as the school shoe pages. Giving the page a way in from somewhere other than the WILDE pages themselves is different work, and it is not one of the itemised changes in the ninety-day proposal, so its hours are not stated separately yet.
Expected result
A customer can reach the page holding all 159 WILDE styles without already being on a WILDE page. No confidence percentage. What this rests on is walking all 108 of your collection pages by hand, not a sample of them.
About two thirds of your orders come from phones. The tile is in the phone row of your homepage, so this is what most of your customers get. On a computer the same row is correct, which is why nobody has seen it.
What a customer on a phone does, and where it ends
We walked every one of your 108 collection pages on 12 August 2026. There is no Wilde School page among them, and the tile on your homepage points at the name of one. The products themselves are fine. On 13 August we tapped that tile on a phone ourselves. It opens a page saying it does not exist, with one button back to the shop and nothing pointing at the school shoes two taps away on your own menu. Three more logos in that row do much the same: Wilde lands on a page of Panda slippers, and Panda Slippers and Aerocushion both land on pages that are not there.
Your phone customers tap a school shoe button on the homepage and land on nothing.
A$18,605Wilde School sales on your own website in the last twelve months. Every Wilde School style listed sells 2.33 times the average for your shop, counted on 12 August 2026. That is turnover per product, not price and not margin.
Give Wilde School a page of its own, point the tile at it, and put the rest of that row right while we are in there. All of it before the January run.
Check it in a minute Open ishoes.com.au on your phone, scroll to the row of brand logos, and tap Wilde School. The row slides, so you may need a second go. Then tap Wilde.
Problem
Phone customers, who place about two thirds of your orders, tap a school shoe tile that leads to a page saying it does not exist.
Solution
Fixing this tile, and the two others in the same row that land the same way, is developer work of the same kind as the other page fixes above. It is not one of the itemised changes in the ninety-day proposal, so its hours are not stated separately yet.
Expected result
The tile leads to a page carrying the 32 Wilde School products, done before the January run. No confidence percentage. What this rests on is tapping the tile ourselves, on a phone, on 13 August 2026.
Just Bee funds its own Google and Meta campaigns for its own store, and it is also promoted on iShoes. So it draws on a constrained iShoes budget while carrying a budget of its own. The brands that budget is meant to serve are Wilde, Wilde School, Aerocushion, Soft Tread and Archies. Two more brands, Zola and Cabello, get no advertising at all.
2025 advertising spend, January to September, one scale from zero
These are the 2025 actual columns from your own planning workbook, one tab per brand, January to September. Nine months, because October to December are empty, so this total cannot be set against a twelve month figure. Just Bee's budget is 3.07 times the iShoes budget. Zola and Cabello have no bar because they have no advertising, and the figure under their name is what they sold on your own website, not what was spent on them.
One brand on your shop is funded three times over on its own account, and still draws on the iShoes budget. Two brands that carry about a sixth of the shop get nothing.
A$39,428Just Bee's own 2025 spend, against A$12,862 for iShoes. On the iShoes own website Just Bee sells A$18,964. That is real revenue and it is not discounted here. Zola and Cabello turn over about A$36,900 between them with no ad budget at all, roughly a sixth of the store. Zola is the only top-six brand shrinking, at about 8% down year on year.
Week one measures what the shared spend costs the brands beside it. If it costs more than it earns, Just Bee is deprioritised on iShoes. Deprioritised, not removed. The products stay on the storefront and keep selling. Only the claim on the paid budget is withdrawn. The offset from freed spend is not measured yet. That is what week one is for.
Check it in a minute Open your own budget workbook, "Meta + Google - Budget Planning (Fujian Only).xlsx". It has three tabs, one per brand: iShoes, Just Bee, Panda Slippers. Read the 2025 actual columns on each, January to September. Then open Shopify Admin, Analytics, sales by vendor, last twelve months, and read Just Bee, Zola and Cabello. Every figure above sits in a system you own.
Problem
A budget meant for five brands is also carrying a brand that already funds its own campaigns, while two brands get nothing.
Solution
Week one is a reading, not a rebuild: what the shared spend costs the brands beside it. Hours for that reading are not itemised in the ninety-day proposal. If it costs more than it earns, Just Bee is deprioritised on iShoes, not removed. The products stay on the storefront and keep selling. Only the claim on the paid budget is withdrawn.
Expected result
A reading of what the shared spend costs the brands beside it, produced inside week one. No confidence percentage: nothing here claims a gain, and nobody should quote one until week one has run, including us.
Every one of them is on a page a customer can open today. They are being found afterwards instead of beforehand, and that is a missing step rather than a missing effort.
Twenty-one things running wrong on your shop right now
What checking would have taken us, against what fixing takes us
Both bars are our hours, not your costs, and both are drawn on one scale with nothing cut
off. Checking is 10.4% of the width of fixing, and that is exactly what you see. Eleven of
those checks take two minutes or less. Nothing in your own figures prices a wrong paragraph,
so we have not put a dollar on this one.
Two more things are worth knowing. The design your shop is running is still named
"iShoes - MAKING UPDATES", and that name is written into every page a customer loads.
And twenty-one of your thirty-three brand and category pages carry a paragraph that describes
a different page. All four Soft Tread pages say "Shop Zola Shoes Online".
Wrong copy and broken links are reaching live selling pages, and they are being found afterwards.
21 live pageswrong today, and every one is a page a customer can open.
Fix the twenty-one, correct the wrong paragraphs, and put a short check in front of publishing that your team runs.
Check it in a minute Shopify admin, Online Store, Themes. Read the name of the live one.
Problem
Wrong copy and broken links are reaching pages a customer can open today, and they are found after they go live rather than before.
Solution
Two tasks. Fixing the twenty-one now is the 23 hours already stated on this page. Putting a short check in front of publishing, so the next twenty-one do not appear, is separate ongoing work, costed in the ninety-day proposal as a catalogue-wide checking pass, eight to twelve hours. The two figures are worked out by different methods, one itemising each of the twenty-one by hand and the other pricing the check as a recurring pass, so they are not added together here.
Expected result
Twenty-one pages corrected, and a check step in place before the next page goes live. No confidence percentage. What this rests on is reading and costing all twenty-one by hand, one at a time.
40,478 words sit in your shop's admin, written and finished. They were live through last September, and they were switched off. Most are old sale copy and are not worth reviving. Twenty-nine of them are usable now.
120 finished articles · live through last September, switched off today
We pulled all 120 out of your admin and read every one. The 29 usable articles run to 11,197 words. Seventeen of them are comfort, fit and shoe care guidance with no expiry date on them. These articles were live on your shop and findable on Google: the Wayback Machine has the whole of School Shoes 101 rendering in January of last year, and the blog busy around it. Today the blog page shows nothing at all, and the article's address comes back not-found. The search places they held went with them.
Finished work you already own was live and was switched off, including the one article written for school shoes.
29 articlesusable now, 11,197 words, and you already own every one of them. Nothing in your figures says what a published article earns, so we have not put a dollar on it. The school shoe one was finished in April 2023.
Revive the 29, with products and collections tagged, images in, and links to the pages they sell. School Shoes 101 first, so it is live and being found well before January.
Check it in a minute The Wayback Machine, your blog's address, any snapshot from last year: the articles, live. Then your own Search Console, the coverage report: the day they dropped out. Then the blog on your own shop from a phone: there is nothing on it.
Problem
Forty thousand words you already own were live through last September and are switched off today, including the one article written for school shoes.
Solution
Reviving the twenty-nine usable articles, and rewriting the four worth rewriting, is publishing work starting in month one and paced across the ninety days: roughly sixty hours in the ninety-day proposal's own costing. Revival means tagging the products and collections each article sells, adding images, and linking the pages, not just switching them back on. School Shoes 101 can run first, ahead of January, or the twenty-nine can run in the order they were written; the sixty hours are the same either way.
Expected result
Twenty-nine articles live on a weekly rhythm, starting in month one and finishing before day ninety. No confidence percentage. Nothing in your figures says what a published article earns, so none is invented here.
Your automated emails, the ones that go out on their own when a customer does something, earn nearly as much as your whole campaign programme, from a fraction of the sends.
Last twelve months to 11 August 2026 · your iShoes account
Per thousand emails sent, your automated emails brought in A$3,241 and your campaigns brought
in A$27.28. Counted a second way, on your own Google Analytics rather than on Omnisend's
figures, automated email still brings in at least four times more per visit.
Customer groups were built in your account. Not one of your last fourteen campaigns used them.
Three sends a week, to the whole list, every time. Nothing in Omnisend joins a group to a send.
That step does not exist in the tool.
The emails that work best go to a few thousand people. The whole list gets everything else.
A$20,407.68came in from 6,296 automated emails. A$27,134.16 came in from 994,739 campaign emails. That is A$3,241 for every thousand automated emails against A$27.28 for every thousand campaign emails.
Put the groups to work, retime the after purchase emails to the first 45 days, and move campaigns to fortnightly sends built on the groups, holding the revenue the whole-list sends bring in today. Each campaign is planned, designed and checked so it lands in your account as a draft. Somebody at iShoes presses send.
Check it in a minute Omnisend, Campaigns. Open the last fourteen and look at the audience field on each one.
Problem
The emails that earn the most reach a few thousand people, and the whole list gets everything else.
Solution
Building segment-aware sending, and strengthening the automations that are already live and earning close to nothing, is one workstream in the ninety-day proposal: twenty to thirty-six hours. The dead automations can be fixed first, or the campaign groups can be put to work first; both sit inside the same hours.
Expected result
The groups already built in your account get used on at least one send, inside month one. No confidence percentage. What this rests on is Omnisend's own count of the last twelve months.
Everything in this report has to be live and being found before your January run, not during it. That is what sets the order, and it is your calendar that sets it.
Ninety days of work, drawn against the calendar
The bar is ninety days from the day you say go, drawn here from today, and the calendar keeps running after it stops. The gap between where the work finishes and where January starts is the whole point of starting now rather than later. Every week the start moves, the bar slides closer to January.
Month one stabilises. The twenty-one defects are fixed, the catalogue is finished top to bottom, the server move is made, and the email and content workflows are run by hand and measured. Week one measures what the shared advertising spend costs the brands beside it. If it costs more than it earns, Just Bee is deprioritised on iShoes. Deprioritised, not removed. The products stay on the storefront and keep selling. Only the claim on the paid budget is withdrawn.
Month two replicates. The same workflows run on Just Bee and Panda Slippers, once those accounts are connected. The social channels come back to life, and email moves to fortnightly sends built on the groups. Month three systematises: what worked by hand becomes the system, and on day ninety we release the agent OS app to you, the work we ran by hand, running under your approval.
The next conversation is where the order gets agreed, and what we start with. Everything above is in your systems today, and every one of the checks in this report takes about a minute. Beside this report sits one live page, plan.html, that takes the ninety days apart: every task, the system it happens in, and the screen you check it on, with the sample brand page, the character-sheet template, and the recordings of the system running our own business.
Problem
Everything above has to be live and being found before January, not during it, and the cost of missing that window is a selling season, not a figure.
Solution
Month one stabilises: the twenty-one defects fixed and the pages and content rebuilt, a mix of itemised and not-yet-itemised work sitting inside the eighty-four and a half hours the ninety-day proposal costs for pages and content, with the email system run by hand for twenty to thirty-six hours. A further eight to twelve hours run across the ninety days checking the rest of the catalogue the same way. Month two replicates the proven workflows to Just Bee and Panda Slippers, once those accounts are connected, and brings the social channels back to life. Month three is the written method and the agent OS app released to you, forty to fifty-four hours. Running it with you, across all three months, is a further six to ten hours. Added up, the five rows run 158.5 to 196.5 hours.
Expected result
Everything in this report live and being found before January, not during it. No confidence percentage. What this rests on is your own calendar, not a sample of other shops.
Prepared by Raava for Fujian Footwear, 13 August 2026, Melbourne (AEST). All figures in Australian dollars. Sales figures are website only unless the page says otherwise. Nothing in this document describes work already carried out.