Open in interactive viewer → charts, metric popovers & call review

Airgain's Enterprise Engine Is Real — Its Promises Are the Problem

A $49M microcap grew enterprise IoT ~50% sequentially, launched an AI antenna platform, and still got 34% shaved off its stock in ninety days. The market is not arguing with the quarter; it is arguing with the conversion record.
AIRG · Earnings Call · 2026-11-10

Airgain's Enterprise Engine Is Real — Its Promises Are the Problem

A story the tape has already stopped paying for

Airgain reported a decent quarter on November 10. Sales of $15.3M rose 9% year over year and 12% sequentially, adjusted EBITDA rose 50% sequentially to $600K, non-GAAP EPS was $0.04, and the fourth quarter is guided to $16.5M at the midpoint — another sequential step up with positive EBITDA. On the surface that is the shape of a business finding its footing. The stock is not buying it. AIRG is down 33.6% over the last 90 trading days, sitting 37.8% below its July 14 peak of $6.18, and 86% below its 2021 high of $28. That is a name where the equity has been repriced for disappointment before the good news arrived. Which means the interesting question is not "did Q3 beat?" It is "why does a company with a genuine growth engine, roughly $7.7M of cash, and a $49M market cap still trade like an option on management's credibility?"

Enterprise IoT is no longer a hope; it is the business

Enterprise sales hit $10M, up $3.3M sequentially and the strongest quarter on record, driven by embedded modems and custom products into energy management, smart buildings, public safety, EV charging and remote monitoring. Total company revenue, which peaked at $19M in 2022Q4, has been rebuilt largely on this line: total revenue of $14M in the June quarter, up 19% sequentially, before the September quarter added another $1.6M. This is the piece of the equity story with real evidence behind it. A strategic IoT customer accelerated a multimillion-dollar order originally slated for March 2027 into September 2026 and added more. Airgain shipped first production units to Coco Robotics, is sampling with three more robotics customers and three drone prospects, and — the new wrinkle — won two design wins in AI data center infrastructure, for remote energy monitoring and rack-level power management, with revenue expected in 2027. That is a deliberate climb up the value chain from Enterprise IoT modems to higher-value applications, and management confirmed a good chunk of Q4's guided growth comes from it. But note the timing. Airgain is pitching AI data center exposure precisely as the public market violently de-rates it — the global AI data center theme is down more than 20% over 30 days across a 59-name cohort spanning chips, power and cooling. Design wins that pay off in 2027 do not offset a sector multiple compressing in 2026.

The conversion problem has a history

Here is where the dossier gets uncomfortable. AirgainConnect — the integrated roof-mounted gateway story — has been "about to inflect" for several quarters. In August the CEO said:

We are also very close to closing a Tier 1 opportunity and we are really wrapping up the last stage which is just the certification that should also really help the second half growth as a whole.

Jacob Suen, President and Chief Executive Officer · 2026-08-05
Three months later, the same Tier 1 public safety opportunity is still open: “We continue to make progress on the Tier 1 public safety opportunity discussed last quarter and are targeting completion by year end.” — Jacob Suen, President and Chief Executive Officer · 2026-11-10 "Second half growth" has quietly become "by year end." The pipeline of 60 Tier 1 and Tier 2 opportunities — roughly 55% public safety, 45% commercial fleet, about half at trial or post-trial — is unchanged in size from the prior call, and four new Tier 2 designs closed in Q3 versus five in Q2. In August the stated ambition was “our goal is to convert at least a third of that every quarter” — Jacob Suen, President and Chief Executive Officer · 2026-08-05; four is not a third of sixty. The product narrative (consolidating multiple data plans into one box, MegaFi 2 and the FirstNet relationship for public safety) is coherent and differentiated. The velocity is not.

Memory: from "no impact" to a real number

Memory was framed as a three-quarter-old risk that finally bit. In May the CFO was categorical: “we had no revenue impact from the shortages, and we had no gross margin impact from the shortages.” — Michael Elbaz, Chief Financial Officer · 2026-05-06 Two quarters later, consumer sales fell $1.4M sequentially to $4.4M as memory shortages disrupted gateway production for cable and mobile operator customers, and the same dynamic is guided to persist: “that same sales dynamic will continue on in Q4 and probably into Q1 as well” — Michael Elbaz, Chief Financial Officer · 2026-11-10 Management flagged recovery from Q2 2027 on the WiFi 7 transition plus MNO programs now holding their regulatory waivers. This is the same recovery language used in the spring about reaching the old $7–8M consumer run rate — a target that has now slipped a full year. Underneath the excuse, though, there is a real distinction: the consumer constraint is volume, not price. Gross margin on those antennas was not hurt; component inflation is being passed through on the IoT side with pricing initiatives. The sales mix shift toward lower-margin custom IoT is what moves gross margin from roughly 44% in H1 to ~42% in H2, and management expects a return to H1 levels within two quarters — a claim worth exactly as much as the last two conversion deadlines. Also worth naming plainly: “Our Q3 number would have been a bit higher than what we have reported” — Michael Elbaz, Chief Financial Officer · 2026-11-10 — a delivery miss on IoT modules and contract-manufacturing capacity, on top of the consumer shortfall. Two supply-driven revenue misses in one quarter is not a narrative problem, it is an execution one.

Airgain RF Intelligence: genuinely new, genuinely unproven

The one item with no precedent in this company's history is Airgain RF Intelligence, an AI-assisted platform intended to make twenty-plus years and a billion-plus shipped antennas of RF engineering expertise self-serviceable for small device developers. The strategic logic is sound: custom antenna engineering has been uneconomic for small programs, and productizing selection from an existing portfolio is a capital-light way to monetize an asset with no incremental design cost. Early evaluation access is targeted for the first half of 2027, with AI-assisted RF engineering as the longer arc. But it is pre-revenue, pre-pilot, and unmonetized — the definition of a 2027-and-beyond option. Coming from a company whose last platform, Lighthouse, was described on a February call as contributing meaningfully in the second half of 2026, the market has learned to discount such launches. Meanwhile the run-rate math is unforgiving: stock-based comp of $895K against $14M of quarterly revenue, comfortably more than the $600K of adjusted EBITDA being celebrated. And receivables at 106% of quarterly revenue, up nearly 19 points year over year, means the enterprise ramp is being financed on the balance sheet — with $7.7M of cash and ATM proceeds of just $200K in the quarter.

What to watch

Airgain has four businesses with four distinct drivers and, for the first time, a coherent growth story in each. The enterprise IoT engine is real, the AI data center adjacency is credible enough to be sampled, and the margin structure (42%-ish gross, $6M of quarterly opex) gives genuine operating leverage on any revenue surprise. What it does not have is a track record of delivering on the words "by year end." The Q4 guide of $15.5–17.5M with positive EBITDA is likely to be met; enterprise demand is strong enough. The re-rating, if it comes, will not hinge on that. It will hinge on whether the Tier 1 public safety deal signs, whether the 60-name pipeline converts faster than four a quarter, and whether Airgain RF Intelligence shows a paying customer before the memory excuse has to be retired for good.