The two things this document says
One. The Australian market is priced far below where you sell. The median listed air purifier is $399, and your cheapest unit at $745 sits above roughly 83% of the models on the market. Your read that price is losing you buyers is correct, and the data says the $595 unit is the better of the two you are planning.
Two, and this is the bigger one. When an AI assistant answers a question about air purifiers, it cites RTINGS 221,183 times, Choice 57,226 times, ProductReview 35,355 times. It cites you once. The site closest to you in authority, airpurifierfirst.com at Domain Rating 31 against your 30, gets 298. Authority is not what separates you from them. Published method is.
PART ONE
The market: what it is worth and how it is priced
203 real listed prices retrieved from 14 Australian retailers and brands on 12 August 2026, deduplicated to 170 distinct models across 29 brands.
What the market is worth
We could not give you a single credible number, so we are not going to invent one. Published estimates for the Australian air purifier market range from AUD $145 million to AUD $621 million, a fourfold spread caused by market research firms defining the category differently and not disclosing how. There is no IBISWorld report for the Australian category, and the ABS does not break it out.
The honest read. The credible cluster sits at AUD $150 million to $400 million, growing 6 to 8% a year. Nobody publishes Australian unit volumes at all. Treat any single figure you see quoted in this category with suspicion, including ours.
How it is priced
$399
median listed price across the market
52%
of models sit at or under $399
76%
of models sit at or under $599
83%
of the market is cheaper than your entry unit
Under $200
29 models, 19.6%
$200 to $399
48 models, 32.4%
$400 to $599
36 models, 24.3%
$600 to $899
17 models, 11.5%
$900 to $1,499
17 models, 11.5%
$1,500 to $2,499
1 model, 0.7%
Your range starts at $745 (E7) and runs to $1,856 (DE20), which places every consumer unit you sell in the top 24% of the market by price.
Two things that keep this honest. Model count is not unit volume, and the difference runs in a specific direction. Kmart, Target, BIG W, Catch and Amazon AU all blocked our crawl, and those are precisely the highest-volume cheap sellers, so the low end is undercounted here. Cheap models also sell far more units per model. True unit volume skews further below $600 than 76% suggests. Separately, the empty $2,500 and up band is an artefact of crawling consumer retail: your commercial range sells B2B where nobody publishes a price. Do not read that gap as an uncontested opportunity.
The $295 and $595 units
| Product | What it walks into | Read |
|---|
| $295 | The most crowded band in the market. 48 models between $200 and $399, dominated by Kmart, Anko, Devanti, Levoit and the retailer house brands, competing almost entirely on price. | Hard. You would be a premium name in a bargain aisle, and margin will be thin. |
| $595 | Materially less crowded. Roughly 24 models in the $500 to $699 zone, and the buyers there are already spending above the median deliberately. | The stronger of the two. Closer to your existing story, and it fills the gap Nick identified: nothing in the range starting with a five. |
The hazard worth naming. Australian manufacture is the wedge that justifies pricing at nearly twice the market median. Imported units cannot carry that claim, so the risk is not margin, it is diluting the one thing that makes the premium defensible. Worth considering a clearly separated sub-brand rather than putting INOVA on both.
Something already in your catalogue. You currently list the Airinum Hale portable at $249, and it sold 115 units for $28,635 last year. There is already an entry-price product on the site and it is not catching the traffic that bounces on price. That points at a visibility problem as much as a range problem, and it is worth understanding before committing to two new SKUs.
Seasonality, and a finding worth acting on
You believe June is your biggest month. Your own data agrees, and June takes 37% of April-to-July revenue in 2025 and 34% in 2026, so it is a real seasonal effect rather than an artefact of the business growing. What is interesting is why.
| 2026 | Search demand
indexed, 100 = average | Organic revenue
indexed, 100 = average | Gap |
|---|
| January | 128, the highest of the year | 88 | −40 |
| February | 107 | 84 | −23 |
| March | 111 | 102 | −9 |
| April | 94 | 95 | +1 |
| May | 85 | 84 | −1 |
| June | 87, the lowest of the year | 146, the highest of the year | +59 |
| July | 88 | 101 | +13 |
Search Console impressions against GA4 organic revenue, indexed within the 2026 window only, so the 143% year-on-year growth cannot distort the comparison. Your own data
June is your best revenue month and your worst demand month. Nothing in the weather explains it: bushfire season in southern Australia has ended by June, and pollen is six months away. What happens on 30 June is the ATO instant asset write-off deadline. Every unit you sell, including the $5,889 commercial machine, sits under the $20,000 threshold, and the asset must be installed and ready for use by 30 June, not merely paid for. A business buyer effectively saves a quarter of the price. It became permanent from 1 July 2026, so this is a repeatable annual play rather than a one-off.
November 2025 shows the same shape even harder, at a revenue index of 207 against demand of 100, which will be Black Friday. Both of your revenue peaks are deadline-driven, not demand-driven.
Which leaves the opportunity. Peak search demand sits in December and January, the southern bushfire season, and your revenue does not follow it. That is a residential demand peak arriving every summer that you are not converting. It is also exactly the audience the organic and AI work in the strategy document is built to reach.
Two caveats, stated plainly. This is 16 months of data, so most calendar months appear once and the direction is far more reliable than the precise magnitudes. And the write-off explanation is an inference, not a measurement. Both are settled cheaply: split your June revenue by residential versus commercial and by price band. If the spike sits in the commercial range, the tax story is confirmed, and the June campaign should lead with the deadline and start in early May. Inference
PART TWO
The teardown: who is winning the answer
Every metric below came from Ahrefs on 12 August 2026, Australian market, or from a direct crawl of the sites named.
Two different fights
Your competitors split cleanly into two groups, and confusing them is the most common mistake in this category. Transaction competitors compete for the sale: Dyson, Philips, Breville, Winix, Ausclimate, the retailers. Answer competitors compete for the recommendation: Choice, RTINGS, ProductReview, HouseFresh, the National Asthma Council. You are losing the second fight badly, and the second fight is the one that increasingly decides the first.
The number that decides the next two years
Ahrefs measures how often a domain is cited inside AI answers. Here is the Australian air purifier category.
| Domain | ChatGPT | Google AI Overviews | Perplexity | Gemini | Domain Rating |
|---|
| rtings.com | 221,183 | 58,128 | 106,078 | 9,266 | 78 |
| choice.com.au | 57,226 | 20,314 | 30,728 | 6,314 | 79 |
| productreview.com.au | 35,355 | 15,064 | 33,641 | 12,157 | 79 |
| dyson.com.au | 3,714 | 1,907 | 962 | 779 | 66 |
| housefresh.com | 386 | 1,076 | 1,211 | 689 | 66 |
| airpurifierfirst.com | 298 | 171 | 670 | 478 | 31 |
| asthma.org.au | 136 | 191 | 156 | 103 | 72 |
| ausclimate.com.au | 27 | 102 | 63 | 63 | 29 |
| inovaairpurifiers.com.au | 1 | 13 | 6 | 5 | 30 |
Sit with the highlighted row. airpurifierfirst.com has a Domain Rating of 31 and 931 referring domains. You have a Domain Rating of 30 and 761. That is the same authority profile within measurement noise. They get 298 ChatGPT citations. You get one. HouseFresh has 566 Australian keywords to your 332, a comparable footprint, and gets 386. You cannot explain this gap with links, budget or brand. The difference is that they publish a test method and you do not.
The answer layer, measured
We built the shared universe of answer-intent terms, the ones containing best, review, which, top, compare, worth it, vs, quietest. In Australia that is 81 keywords carrying 23,470 monthly searches.
| Site | Terms it ranks for | Volume covered | Volume held in the top three |
|---|
| choice.com.au | 47 of 81 | 17,730 | 16,530 |
| rtings.com | 37 of 81 | 11,370 | substantial |
| housefresh.com | 19 of 81 | 6,450 | moderate |
| airpurifierfirst.com | 11 of 81 | 4,700 | limited |
| INOVA | 8 of 81 | 3,680 | 680 |
You do not rank at all for 73 of the 81 terms, worth 19,790 monthly searches. The six answer terms you do hold in the top three are all narrow and specification-adjacent: best air purifier for mould Australia, best air purifier for smoke Australia, best HEPA air purifier for covid Australia. Good, defensible, and worth protecting. But together they are 680 monthly searches against Choice's 16,530. You own roughly four percent of the answer layer you should be contesting.
The twelve terms that decide who AI recommends
These are the terms where three or more independent review sites all rank. When three independent domains agree, that is the pattern a language model reads as consensus, and those are the pages it draws from. Together they carry 6,800 monthly Australian searches. You appear on exactly one of them, at position 15.
| Volume | Term | You | Who owns it |
|---|
| 2,000 | best air purifier | 15 | Choice #1, RTINGS #1 |
| 1,300 | best air purifiers for allergies | none | HouseFresh #1, Choice #2 |
| 700 | best air purifiers | none | Choice #1, RTINGS #1 |
| 500 | do air purifiers remove dust | none | Choice #4, HouseFresh #5, RTINGS #8 |
| 450 | do air purifiers help with dust | none | Choice #1, airpurifierfirst #5 |
| 350 | air purifier reviews | none | Choice #1, RTINGS #1, HouseFresh #1 |
| 300 | best home air purifiers | none | RTINGS #1, HouseFresh #3 |
| 300 | best pet air purifier | none | RTINGS #1, HouseFresh #1 |
| 250 | best air purifiers for bedrooms | none | airpurifierfirst #2, Choice #3 |
| 250 | best dust air purifier | none | Choice #1, HouseFresh #3 |
| 200 | best air purifier home | none | RTINGS #1, Choice #3 |
| 200 | reviews for air purifiers | none | Choice #1, RTINGS #1, HouseFresh #1 |
Difficulty across this tier runs from 0 to 33. "Are air purifiers worth it" is difficulty 0. "Do air purifiers help with dust" is 1. These are not hard terms. They are simply unclaimed by anyone Australian who actually tests. Ahrefs
The Choice problem
Nick's account of Choice matters commercially, so we have separated what he told us from what we independently verified. That separation is deliberate, because the verified part is stronger.
| Nick's account | What we could verify |
|---|
| Choice charges around $20,000 a year to licence its logo, and brands that pay rank top | Not verified, and we would not publish it. The fee is not published anywhere. There is also no "CHOICE Approved" programme, it is CHOICE Recommended. |
| Choice's testing favours airflow over filtration quality | Supported by their own published material. Choice publishes a 70/30 and 50/40/10 weighting, and states in its own words that testing every unit in the same room means "bigger models usually perform better". |
| Choice would not answer questions about chamber size and equipment | Consistent with what they publish. Choice does not publish chamber volume, any named test standard, particle sizes, test duration, equipment or lab name, and does not test air purifiers in its own facility. |
The stronger, unrebuttable version of the argument. Choice's own FY2024-25 annual review discloses $1.3 million of revenue, 6.1% of its $24.2 million total, from manufacturers licensing its endorsement. It publishes no standalone conflict-of-interest policy and offers no appeals process for a disputed score. Every word of that is Choice's own published material. It makes the point without asserting anything you would have to defend.
And the finding that actually matters. Choice's public product pages for your E20 and E8 both show "Claimed CADR dust removal: Not stated" and "Claimed bacteria and virus removal: 0%". Their score is CADR-weighted by construction, with dust alone worth 35% of the total. You are being scored on a metric you have declined to declare, and the 0% is almost certainly correct regulatory caution rendering as a zero. That is worse than a bad score, because it is a blank that anyone can see, including the crawlers. Publishing an independently measured CADR and supplying it to Choice is the single highest-value fix in this document.
Before you publish anything critical of Choice, read this. The strongest argument against Choice is that it takes money from manufacturers for an endorsement. That argument cuts both ways, and you need to know exactly how far before you make it. The National Asthma Council's own partner page states that Sensitive Choice "is not a standards-based certification program. It is a marketing/certification program" requiring "payment of an annual licence fee… not less than $6,000 in Australia". The Council does not test: applicants supply their own independent test reports and a volunteer panel reviews the file. And the Council states that participation "does not represent that the product or service is the best or better than others in a category". That fee is disclosed only on the partner-recruitment page, not on any consumer-facing page explaining what the butterfly means.
What that means practically. Attack Choice on pay-for-endorsement and a competitor can point at your butterfly and run the identical argument. The defensible position is the one that does not depend on anyone else's integrity: publish your method and your numbers, and invite comparison. Criticise Choice on what they do not publish, which is chamber volume, test standard, particle sizes, duration, equipment and lab, because that critique is about transparency and you can answer it about yourself in the affirmative.
Five things to do, in order
1
Publish your own test methodology and data
Highest
This is the whole strategy. RTINGS built 221,183 citations on a versioned public test bench. HouseFresh built 386 on a published chamber volume, named instrumentation, particle-resolved channels and downloadable data. You have the lab and the machines and publish none of it. Build a permanent, versioned methodology hub: chamber volume, named particle counter, efficiency at 0.3 / 0.5 / 1 / 2.5 / 10 microns, decay curves, measured CADR, noise in dB at each speed, wattage, raw data, a method version number and a named responsible engineer. Say plainly that INOVA funds the work and name any independent verification. The "marking your own homework" objection is beaten with disclosure, not by publishing less.
And there is a vacancy here that makes this unusually winnable. We checked whether an Australian performance standard already fills this gap, and could not identify one. AS 4260 covers HEPA filter elements. AS 16890 covers HVAC ventilation filters. AS/NZS 60335.2.65 covers electrical safety. None of them rates how well a finished purifier cleans a room, and the international standard that does, IEC 63086, has no Australian or New Zealand adoption we could find. Whoever publishes a credible Australian test method first effectively defines how this category gets compared.
2
Claim the twelve consensus terms with a real Australian comparison hub
Highest
Choice concentrates 35 of its 60 air purifier keywords on one URL. RTINGS concentrates 29 of 40 on one URL. Both won this market with a single canonical page plus use-case spokes. Your rankings are scattered across the homepage and collections with no canonical answer page at all. Build one, on your own published test data, honest about competitor machines including where they beat you, with spokes for allergies, dust, mould, pets, bedrooms, smoke and asthma. RTINGS is American, HouseFresh is American, airpurifierfirst is American and Choice is paywalled. There is currently no Australian, test-based air purifier comparison resource at all. That vacancy is the biggest structural opportunity here.
3
Occupy productreview.com.au deliberately
Fastest payback
ProductReview is Domain Rating 79 with 944,438 Australian organic visits and 35,355 ChatGPT citations, the largest Australian answer surface in this analysis, and you already have listings on it. Systematically grow genuine review volume, recency and depth, and answer every question in the Q&A sections properly and technically. Not review-gating, which breaches consumer law. Just asking every satisfied customer, every time, through a proper post-purchase sequence. Every review sits inside a document the models already trust, so there is no crawl-and-trust delay. Weeks, not quarters.
4
Turn Sensitive Choice into an asthma.org.au citation
Best value per unit of effort
asthma.org.au holds only 1,929 Australian keywords, yet two blog posts own the entire asthma and allergy segment, including position 1 for asthma air purifier and position 9 for "best air purifier australia" at 2,700 a month. You rank for none of it, and you are a Sensitive Choice licensee of that exact organisation. Approach them as a licensee with something to offer: particle-resolved capture data on the allergens that matter clinically, dust mite, pollen, cat and dog dander, mould spore. Aim for co-authored or clinically reviewed content. Use the mark as evidence of clinical acceptability, never as a performance claim. The Council's own published wording is that participation "does not represent that the product or service is the best or better than others in a category", and that the programme "is not a standards-based certification program". Anything you say about the butterfly needs to survive being read next to those sentences.
5
Close the Ausclimate gap and expand the commercial pages
Cheapest and most certain
Ausclimate sits at Domain Rating 29 with 763 referring domains, effectively identical to your 30 and 761, and pulls 17,004 Australian organic visits from 499 keywords against your 3,949 from 332. Same authority, four times the traffic. The difference is content coverage and page architecture, not links. Meanwhile your two flagship product pages run to 7,800 words each and carry no FAQ schema, so the best raw material on the site is structurally invisible as an answer. Add FAQ and Product schema across every substantive page, then expand the commercial, healthcare and education pages to the depth Ausclimate has built on the consumer side. There are 61 keywords worth 12,530 monthly searches where you rank and no review site competes at all, concentrated in exactly that commercial block. Lower volume, far higher order value. This one does not depend on the answer-layer strategy landing, which is why it runs in parallel rather than after.
The line every claim has to stay behind
Publishing test data is the right strategy, and it raises the stakes on getting the wording right. Two rules govern this, and they are not the same rule.
| Regulator | What triggers it | What it means for you |
|---|
| ACCC | Any performance claim at all. Their published rule is that claims "must be accurate, truthful and based on reasonable grounds" and "a business must be able to prove any claim they advertise". | Publishing your method strengthens your position here. Unsubstantiated claims are the exposure, and evidence is the defence. |
| TGA | Claiming the device kills or inactivates viruses or bacteria, or that it treats or prevents asthma or allergies. | This is the sharp edge, and you have a Virus Series. Therapeutic claims can convert an ordinary appliance into a regulated therapeutic good requiring ARTG listing before supply. |
The precedent worth knowing. In 2021 the Federal Court penalised Lorna Jane $5 million after it admitted representing there was a scientific basis for anti-virus claims when no such testing had been carried out. The lesson is not that claims are dangerous. It is that claims without evidence are dangerous, which is precisely the problem publishing your test data solves. We found no ACCC or TGA enforcement action against an air purifier marketer, so this is a boundary to respect rather than a live fire.
Practical rule for the engagement: performance claims get expressed as measured filtration performance under a stated method, never as a health outcome. Everything in a regulated category gets referenced and gets your approval before it goes live. Verify TGA wording against tga.gov.au before publication
What we could not establish
- Choice's actual scores for your machines are paywalled. We verified only the public comparison fields. The poor-rating account is relayed from you, not independently confirmed.
- That no Australian performance standard exists, as a proven negative. We identified the three AU/NZ standards that touch this space and confirmed none of them rates appliance performance, and we found no AU/NZ adoption of IEC 63086. But Standards Australia's catalogue could not be searched authoritatively, so the honest claim is "none could be identified", not "none exists".
- Australian unit volumes by price band. Nobody publishes them. Our distribution counts models, not units sold.
- Samsung's Australian pricing was excluded because their site served prices in GBP against mismatched model codes, and we would rather drop a brand than publish a number we do not trust.
- Australia-specific search seasonality from Google Trends. We used your Search Console impressions instead, which is better data for your purposes, but Trends would confirm the category-wide pattern.