Be the cabin brand
AI recommends in New Zealand.

1. Executive summary

• Your organic traffic is category demand, not brand demand, and that is rare. Most brands we audit earn their organic revenue from people who already know them. Yours is the opposite, helped by a smart naming decision: SheShed is both your brand and the thing people search for, so even someone typing "she shed" with no brand in mind can land on you. Counting only true navigational searches as branded, roughly 9 in 10 organic clicks are category demand, and you already take 24.5% of the winnable click pool. You are winning strangers, which is exactly what a search strategy is for. GSC
• Nobody can currently prove which channel sells cabins. 93 of your 113 tracked online orders, $1.87M of the $2.22M, sit in GA4 as Unassigned. The purchase tracking fires, but the attribution is lost by the time it does. Getting accurate data here, including real conversion rates by channel, would be genuinely valuable: it sharpens every marketing decision you make, and it firms up every dollar figure in this report. GA4
• AI is already in your funnel, and the site can be made much readier for it. ChatGPT sent you 373 visits this year with zero optimisation, and Google AI Overviews already fire on "she shed" and "sleepout". One quiet drag: in Google's mobile lab test, which simulates a mid-range phone on a slow connection, the homepage's main content lands at 19.9 seconds. The site feels quick on a good connection, and that lab figure is still the one Google measures rankings against, which makes it one of the cheapest wins on the board. GA4 Lighthouse

So the whole opportunity is one sentence: you hold 24.5% of the non-branded organic clicks in your category, and every extra 1% is worth about $7,800 a year. Your branded traffic is not part of this and does not change.

Built on your own numbers, with one honest caveat up front: because of the attribution gap, we do not yet have a measured organic conversion rate for your site, so this model does not pretend to one. Instead it uses what is solid: 113 orders on $2.22M gives an average order of $19,655, and across all 247,816 visits that makes a visit worth $8.96, however it arrived. We value an organic click at that blended figure rather than guessing a channel rate. Treat the dollar columns as directional; the moment real conversion data is available, we re-run the model on measured numbers and these figures firm up in both directions. Every column below is non-branded, and your branded traffic sits on top of all of it, unchanged. Modelled, plus or minus 40%

At your average order of $19,655, the mid case is about seven extra cabins a year, every year, compounding as the content keeps ranking.

Triangulated NZ search volume per category, August 2026. The full keyword, prompt and cluster research behind every number here lives in The data tab. GKP + Ahrefs + SEMrush

2. How AI-ready you are today: the AEO scorecard

We score every brand out of 100 against our AEO Maturity Model: four pillars (Technical, Content, Authority, Measurement), each out of 25. Certified is 80 out of 100, Professional on all four, the standard we build clients to. You are at 40 today, and unusually, your content is the strength and the plumbing is the problem.

The detailed findings, how the scoring works, and where every number comes from are in the appendix.

3. The market you are missing: search and AI

A cabin is a considered purchase. Your buyers spend weeks researching consent rules, foundations, timber, power and cost before they get anywhere near a checkout, and they now ask those questions in two places, a search box and an AI chat. The demand is deep, most of your direct competitors are tiny, and the AI layer has already started answering.

Search volume triangulated across Google Keyword Planner, SEMrush and Ahrefs (median of the sources returning a figure), NZ. Your rank is your current average Google position from Search Console. The pattern: bright spots where you have a dedicated page (pool house, alpine cabins), page two to five everywhere else, including your namesake term. Page two is where 21,282 clicks came from; the top three is where the other 60,000 live. The full 112-term triangulation is in The data tab, including the terms your own Search Console proved that the tools understate, like garden rooms and alpine pods, marked "GSC-proven". GKP + Ahrefs + SEMrush GSC

Zoom out and the full category is about 36,400 searches a month once the long tail is counted, of which roughly 7,250 a month are winnable organic clicks, plus the AI-chat questions counted separately. In the live results the pattern is simple: you win log cabins and kitset tiny homes, you are contested on she sheds, and you are missing from page one on kitset cabins and sleepouts, the highest-intent terms in the book, where a field of much smaller specialists currently holds the ground. The full 130-question library, cluster maths and territory plan live in the appendix and The data tab.

4. The opportunity, in dollars

The table in the executive summary is built from four numbers, all of them yours, and all of them non-branded. Here is the chain in full.

Roughly one search in five survives ads, shopping carousels and AI answers as a winnable organic click, and you currently take one click in four of those. Modelled

Each visit to your site is worth $8.96, because 247,816 visits produced $2.22M in orders. That is the figure every case in the table multiplies. It is a blended site-wide value, used because no per-channel conversion rate can currently be trusted; it firms up the day real conversion data is connected.

That is the whole model: win more of the 87,000 clicks.

The mid case arriving, on non-branded revenue: certified inside the six months, the system handed to you at month 6, all of it live by month 12 (~10% live by month 3, ~35% by month 6, ~70% by month 9). Hover any point for the maths. Modelled, directional

Two levers, both measured. Click-through on the impressions you already get: you appear on 525,595 non-branded impressions a year but convert them at 1.21%, because your average position is 15. Moving the terms you already rank for from page two into the top three multiplies the clicks without a single new impression. And more impressions, as the consent, cost and comparison questions you have no ranking page for start to rank. The conservative case assumes only the first lever really lands; the high case assumes both do.

Every figure here is non-branded. Roughly two-thirds of each case is organic search and one-third is AI search, modelled from the prompt library demand in the appendix. Total organic run-rate is your branded traffic, unchanged, plus the non-branded column in the executive summary table.

5. The six-month transformation

A done-for-you AI and Search transformation over six months, built and run on our side so there is nothing for your team to learn or maintain. It moves through three phases: foundations (the mobile speed scores, the schema clean-up, the attribution rebuild so every order traces to a channel, your baseline), owning your territories (the consent and cost answer hub, the sleepout and kitset cabin hubs, and reshaping your existing 160 posts into the answers engines cite, with new production starting once everything is calibrated to your voice), and authority and compounding (extending the authority lead you already hold over every direct competitor into the citations AI trusts). We move into each phase as your site is ready for it rather than on a stopwatch. Behind it, we are building the platform you will own, built for ecommerce brands: by the end of your six months it is a system your team can log into and run. Everything is yours to keep: the foundations, the content, the links and your AI-Ready certification.

The full breakdown, deliverables, your time commitment and investment live in The offer tab.

Traffic, orders, revenue and rankings are from your own Google Analytics and Search Console, last 12 months to 16 August 2026, NZD. Search volumes triangulated across Google Keyword Planner, SEMrush and Ahrefs (NZ), August 2026. Competitor Domain Ratings from Ahrefs. AEO score from our AEO Maturity Model against a live crawl and a mobile Lighthouse run. Revenue figures are directional, modelled from your own order value and traffic, and are not a guarantee of results. Investment details are set out separately.

Appendix A: your channels and current position

You have built a real brand in a category you named yourselves into: $2.22M of tracked online orders in 12 months, a 17-product kitset range from $11,995 to $76,990, NZ-wide delivery and a content library your competitors have not come close to. The foundations are good. This is growth on a proven base, not a rescue.

GA4, last 12 months to 16 August 2026, NZD. Enquiries are GA4 key events (forms, calls, downloads). The Unassigned row is the finding: the purchase event fires, but by the time it does, GA4 no longer knows where the buyer came from, so $1.87M of revenue credits no channel at all. Until that is fixed, no channel figure in this table can be fully trusted, which is exactly why the model values a visit at the blended $8.96 rather than any per-channel rate. GA4

Non-branded search is 94% of your impressions and 77% of your clicks, but the click-through tells the real story: branded queries convert at 6.37% and non-branded at just 1.21%, because on category terms you sit at position 15, not the top three. The visibility is there; the position is not. GSC

Appendix B: the scorecard in detail

Here is what sits behind the score, and who fixes each piece. Unusually for our audits, the content engine already exists. The work is speed, schema and measurement, then shaping what you have written into answers.

The single highest-leverage fix is speed. To be clear about what the numbers mean: the site loads promptly on a good connection, and the figures above come from Google's Lighthouse lab test, which simulates a mid-range phone on a slow connection. That simulated test is the one Google's ranking systems draw on, so the gap between how the site feels and how it scores is exactly the gap worth closing. The second fix is schema: your pages still carry a previous agency's Australian template, complete with their logo and a self-awarded five-star rating, which is exactly the kind of signal that erodes machine trust.

Appendix C: the demand detail

The triangulated core (26,985 a month across 112 terms) grows to a full category of about 36,400 searches a month once the long tail is counted. Strip out ads, shopping carousels, AI Overviews and pure-research queries, and roughly one search in five survives as a winnable organic click, a pool of about 7,100 a month. AI prompts are counted separately so nothing is double-counted. Modelled

One check we always run: the demand model is validated against your own Search Console and GA4, never just the tools. For you that check added garden rooms, summer houses, alpine pods and workshop demand the tools understate, all already ranking in your account. It also proved the consent territory is far bigger than any tool suggests: your own outbuilding-versus-dwelling explainer earned roughly 22,000 impressions last year on definitional consent queries alone (dwelling, outbuilding, and their variants), sitting at positions 4 to 11. Those queries never enter the click model, because a definition rarely earns a click, but they are precisely the citations AI engines lift, and that page is the seed the consent hub grows from. GSC + GA4

We built a library of 130 real questions Kiwis ask AI engines about your category, from Google's "People Also Ask", the consent rules every buyer hits, and the head-term intent map. Here is a sample, with roughly how often each comes up a month.

A sample of the 130-question library, all of it in The data tab with per-question volume, priority and ownability. Volumes are directional floors (head-term anchor plus fan-out uplift). 124 of the 130 are questions you can uniquely or credibly own. Directional

And here is who the engines surface today, cluster by cluster, versus where you show up.

Live Google results, August 2026. The consent row is the one to stare at: the question behind almost every cabin purchase is currently answered by councils, which cannot recommend a product. Live results

Appendix D: where to play, your content territories

The same demand, grouped into the territories a plan executes against, each scored on how much of it you can realistically own. Priority weighs demand, whether AI is live, and whether you can win it uniquely.

Priority is a 1 to 5 score, where 5 means lead with it now; it is not a count of pieces. Content requirement: roughly 55 pieces to own the core (consents, sleepouts, kitset cabins, she sheds), and 100 to 140 over the six months for full category coverage. Unlike most brands we audit, much of this is reshaping, not writing from scratch: your 160 posts and 28 use-case pages are the raw material, restructured into answer format, retyped in schema, and interlinked with the shop. Every piece also routes to your catalogue and buying-guide downloads, which drove roughly 1,000 enquiries in the last year and are your measured conversion layer. The full cluster maths is in The data tab. Cluster model

Appendix E: who you are up against, and the authority play

Ahrefs Domain Rating, August 2026. The competitors who outrank you on kitset and sleepout terms are doing it with less authority than you have, purely on page relevance. That is the most fixable gap in search. Ahrefs

Our honest sequencing: the foundations and content phases get you cited on your own answers and lift you from page two into the top three. Authority, the digital PR, round-up placements and community presence, is the engine that runs alongside and then compounds for years after. It is where most of the long-term share-of-voice sits.

Appendix F: where the numbers come from

Measurement runs in our own platform, built for ecommerce brands, so the dashboard, the data and the method are yours to keep, not rented from a third-party tool. The full research behind every table here is in The data tab. Method