Search for RFID market 2026 chip shipments data and you will find page after supplier page describing uninterrupted growth. The industry’s own trade body reported a figure that complicates that story, and anyone about to sign a multi-year tag contract should see it before the growth chart.
The number most RFID supplier blogs leave out
The RAIN Alliance, the trade body for UHF RFID, reported that 42.7 billion RAIN tag chips shipped in 2025, and reported it as a fall against the year before. On the trade body’s own numbers, 2025 was a down year.
What the three-year series actually looks like
The Alliance publishes a tag-chip total once a year, each one in a release of its own. Put the last three side by side and the shape tells you more than any single headline does.
| Year | RAIN tag chips shipped | Alliance release | Reporting members |
|---|---|---|---|
| 2023 | 44.8 billion, which the Alliance put at 32% above 2022 | Body dateline 4 March 2024 | Four, described as the leading tag manufacturers and all Alliance members |
| 2024 | 52.8 billion | Body dateline 26 February 2025 | Five: the four listed for 2025, plus Shanghai Fudan Microelectronics Group |
| 2025 | 42.7 billion | Published 2 April 2026 | Four: EM Microelectronic, Impinj, NXP and Shanghai Quanray Electronics |
How the Alliance describes the source matters as much as the totals. The 2025 and 2024 releases attribute the figure to “its leading tag manufacturer members”; the 2023 release words it differently, crediting the four leading tag manufacturers and noting that all of them are Alliance members, and it names the Chinese contributor as Quanray rather than Shanghai Quanray Electronics. Either way, this is a compilation from a handful of reporters, not a census of every RAIN chip made anywhere on earth — and the naming drift is worth knowing if you go searching the releases to check us.
2025 fell against 2024 and still finished above 2023. That is the point of printing the series, and it is neither of the two stories you will be sold. It is not a boom: the record year was 2024, and 2025 did not match it. It is not a collapse either: the down year still shipped more chips than the year before the record. A record followed by a correction that lands above the pre-record level is an ordinary shape for a component market working a supply shock out of the channel — and it is a far more useful thing for a buyer to know than either extreme.
Two things to understand before you reuse the series anywhere.
No single Alliance page carries all three figures. The 2 April 2026 release gives 42.7 billion and nothing else; it refers to the previous year only as a record one, without a number and without a percentage. The 52.8 billion and the 44.8 billion each come from their own earlier release. If someone hands you a tidy three-year series with one citation attached, they have not assembled it from the releases. Dates need the same care: the Alliance’s own pages for the 2023 release do not all carry the same date, which is why we cite the dateline printed in the release body, 4 March 2024. If a figure’s date is load-bearing in a tender document or a board paper, archive the page you actually read.
We are not printing a year-on-year percentage, and that is deliberate. The 2024 total was reported by five member companies; the 2025 total by four, with Shanghai Fudan Microelectronics Group absent from the later list and no published explanation for the change. The Alliance gives no percentage for 2025, and does not say whether it restated 2024 onto a comparable four-company basis. So the obvious arithmetic spans a change in who is being counted. That observation is ours, from comparing the two member lists — not an Alliance statement — and how much of the fall is a real contraction versus one fewer reporter cannot be settled from what has been published. The direction is well supported. A decimal-place delta is not.
For the same reason we do not print a 2022 base figure. We have not found one published by the Alliance as an absolute number; it appears only to be derivable from growth statements, so it stays out.
One further trap, because it is exactly the kind this post exists to flag: VDC Research’s own figure for 2023 is 45.5 billion, not 44.8 billion, and the Alliance says plainly that the VDC number is the broader of the two because it also counts manufacturers shipping smaller quantities. They are two different measurements of the same year. Quoting both in one paragraph without saying so is how a sourcing document ends up contradicting itself.
Identium is a tag and reader manufacturer in New Delhi, not a chipmaker. We have no visibility into any member’s shipment reporting, and nothing above comes from anywhere but the Alliance’s published releases. The chips we actually buy and build around are covered in our guide to the Impinj and NXP chip families.
When this post first went up we declined to print the prior-year figures at all, on the grounds that we had not opened a primary source for them and a post with a whole section about unsourceable numbers should not smuggle one into its own opening. That was the right call then and it is out of date now: the Alliance has published each year’s total in a release of its own, so the series above is drawn from those releases, each dated above. The scepticism has not moved — it has been applied to the right target, which is the percentage rather than the figures. Check the current numbers with the RAIN Alliance directly before you quote them anywhere.
Two caveats on every figure above. RAIN chips are UHF chips, not all RFID: HF, NFC and LF sit outside the count, so none of these totals is a headline about the whole industry. And each total is compiled from the Alliance’s leading tag manufacturer members rather than from a census of every chip made anywhere, so the signal is the direction of travel rather than an absolute. A chip shipped is also a chip sold into the channel, not a tag applied to a product.
Why 2025 dipped, in the RAIN Alliance’s reported explanation
The Alliance did not present this as demand falling away. Its reported explanation named three things: semiconductor inventory cycles, tariff uncertainty and retail destocking.
All three are distribution-layer effects rather than end-user ones. That is what a destocking year looks like from the fab: converters who stocked heavily through the shortage years draw inventory down instead of reordering, retailers run their own stock down, and buyers facing uncertain duty treatment hold off committing to volume. A year in which fewer chips left the fab is not automatically a year in which fewer tags went onto products.
The four-year trend is still up — holding both facts at once
The same communication made the counterpoint, and it deserves equal weight: new deployments continued through 2025, and the market is reported to be up roughly 50% over four years.
Both things are true, and the series above is what holding both looks like on paper: a record, then a correction that still cleared 2023. A single-year fall inside a four-year climb is an ordinary shape for a component market coming off a supply shock. The forecast actually on the record is the one the Alliance published itself on 3 June 2024, drawn from VDC Research: 115 billion units by 2028, at year-on-year growth of 20.4%. Read it for what it is — a forecaster’s projection carried in a trade-body release, made nearly two years before the 2025 figure existed — not as a floor. And if you are shown an Alliance projection of 64 billion chips for 2025, ask which release it is in, because we could not find that figure in any Alliance publication we opened.
We make no claim about 2026 shipments. The year is not over, and anyone telling you how it ends is guessing.
The figures we could not verify, and why they are not in this post
This section is the reason the page exists. Three numbers appear on nearly every RFID supplier blog. We could not trace any of them to a primary source we can open and check, so they are not in this post.
| Widely quoted claim | Where it usually traces back to | Why it is not here |
|---|---|---|
| The RFID market is worth about USD 19 billion | Press-release wire summaries of paid analyst reports | The methodology and scope sit behind a paywall. Hardware only, or hardware plus software and services? The number changes completely depending on the answer |
| The market grows at about 11.5% a year | Consolidated analyst estimates, republished supplier to supplier | Firms define the market differently, so a blended CAGR cannot be checked without the model underneath it |
| Inlays now cost under USD 0.04 | Trade commentary and vendor marketing | Real inlay price depends on chip, antenna, substrate, volume and converting. A single global figure is not something a buyer can hold a supplier to |
None of those figures is necessarily wrong. We simply cannot show you where they come from, and a number you cannot source is a number you cannot negotiate against. Our RFID tag price guide for India works the other way round, from specification to an indicative band. The one figure we state as a firm anchor rather than an indicative range is RFID wristbands from about Rs 15 per unit; everything else we quote against a written spec.
The other 2026 story: the supply side consolidated
While volume dipped, the ownership map changed. One published deal list for the first half of 2026 is RFID News’ acquisition tracker — a single aggregator’s count rather than an audited tally, so treat it as one source rather than the record. The largest transaction on that list is Brady’s acquisition of Honeywell’s Productivity Solutions and Services business, reported at approximately USD 1.4 billion in April 2026, which is independently reported elsewhere.
Identiv’s exit from RFID manufacturing, and what it signals
The clearest single event is documented in SEC filings. On 15 September 2026, Trackonomy closed its acquisition of Identiv’s IoT business — the RFID transponder and inlay manufacturing operation, with the transferring assets reported to include a German R&D centre and a Thai subsidiary. The structure ran in both directions: as reported, Identiv contributed USD 25 million in cash in exchange for USD 50 million of Trackonomy preferred equity. The listed remainder was renamed INVE Technologies and is pivoting to compliance software, with James Greenwell appointed interim CEO effective 21 September 2026. The filings are public — read them rather than take our summary.
Nothing here says any buyer is stranded. The manufacturing assets moved to a new owner rather than shutting, and how that owner serves existing customers is not ours to characterise. The signal for a buyer is narrower and more useful than any drama: a long-established Western transponder and inlay supplier is no longer in that business under that name. If your contract, part number and tooling were attached to a legal entity rather than to a written specification, that is the kind of event that tests your paperwork.
Where the money went in H1 2026: software and applications, not inlays
The tracker’s own conclusion about the first half of 2026 is that capital went to companies combining tags with software rather than to commodity inlay makers. INVE’s pivot to compliance software is that same pattern seen from the inside. It is one aggregator’s read of a partial list, so hold the analysis loosely — but the practical consequence for a buyer does not depend on it being exactly right. Ownership of a tag line can change mid-contract, as 2026 has shown, and a contract written as though it cannot is a contract with a clause missing.
What a buyer should check about supplier continuity before signing
Use this whoever you buy from. It is written to be useful even if the answers send you to one of our competitors.
| Ask | Why it matters | What a good answer looks like |
|---|---|---|
| Who owns the line our tags are made on? | Many suppliers subcontract. A subcontracted line can be reassigned without telling you | A named site the supplier operates, or an openly named contract manufacturer |
| Where are they physically made? | ”Asia” is not an answer you can audit or declare | A site address, and willingness to let you or an agent inspect |
| Does our part number survive a change of ownership? | Part numbers are entity-specific; specifications are not | The part number tied to a spec sheet naming chip, antenna, substrate, size, encoding and test criteria |
| Who owns the tooling we paid for? | A mould you funded can otherwise sit as the seller’s asset | Your ownership named in the contract, with a release condition |
| Is a repeat-order rate agreed, and for how long? | Reorders at “current price” are where budgets break | A stated rate with a validity window and a defined revision trigger |
| Which chip is inside, and is an alternate approved? | A chip substitution can change read performance without changing the part number | The chip named, with any alternate qualified and tested before it ships |
| Is there a second qualified source? | Single-sourcing a part that halts operations is a risk decision, not a default | A second supplier holding your spec, even at low volume |
Our guide to choosing an RFID tag manufacturer covers the audit questions in more depth, and the India versus China sourcing comparison covers what a genuine second source costs to qualify. For private-label buyers, the OEM and ODM guide sets out which clauses get missed.
What all this means if you are budgeting an RFID project in India
Three practical conclusions.
Do not budget from a global average. Build the number from your own specification — tag family, environment, read distance and annual volume — then quote it. Our ROI and payback guide shows which line items usually decide whether a project pays back, and they are rarely the tag price.
Write continuity into the contract, not into the relationship. Specification sheet, tooling ownership, repeat-order rate, approved alternate chip. If you are importing, the MOQ, lead time and Incoterms guide covers the commercial terms that sit alongside them.
Expect regulation to move on its own timetable, not the shipment curve’s. Item-level identification requirements such as the EU’s digital product passport advance regardless of a soft chip year — though reporting as of 2026 points at a textile delegated act around 2027 rather than imminently, so treat every date you are quoted as reported until the European Commission’s own text is final. Our DPP briefing for Indian exporters covers what that means for a factory tagging at source.
Identium has built RFID hardware from New Delhi since 2010, incorporated 2018, with converting, ultrasonic sealing and encoding in-house and custom hard tags typically turned around in 5–10 days once the design is frozen. We are BIS certified, and our UHF readers and antennas are WPC-approved for India’s licence-exempt UHF RFID band. On the regulation rather than on our paperwork: that band is 865–868 MHz, set by G.S.R. 853(E) of 10 December 2021, made expressly in supersession of the 2005 rules that had allowed 865–867 MHz. Rule 1(3) grandfathers equipment type-approved under those older rules for its service life, which is why a band figure and the date on any supplier’s approval are two separate questions. Ask us, or anyone, for the approval document itself. We do not hold CE, FCC, RoHS, REACH or ISO 9001 — if your market needs those marks, weigh it as a real gap and ask every supplier the same question.
If you want a part number tied to a written specification rather than a catalogue photograph, tell us what you are tagging and we will quote against the spec.