Shopping in the LPWAN Market? Choose an LPWAN That Will Survive.

If wireless history is any indication, there will most likely be a single “big” winner for LPWAN.


I recently came across a new report, Survival of the Timeliest-Can Unlicensed LPWA beat NB-IoT?, with conclusions about the LPWAN market. Two things caught my eye:

1. There can be only one LPWAN winner in the unlicensed spectrum.

If wireless history is a guide, there will most likely be a single “big” winner for low power, wide area narrowband connectivity. Same as it was for local area wireless broadband (WiFi) or personal area broadband (Bluetooth).

The licensed vs. unlicensed distinction, for me, is driven by today’s wireless carriers who present themselves as a serious potential distribution channel for low power IoT connectivity, which is very much TBD.

Moreover, some of the larger carriers are skipping an LTE-based LPWAN option and embracing non-LTE LPWAN’s instead. Which frequency band(s) the LPWAN operates within — licensed or unlicensed — is immaterial to whether it is the big winner.

2. There is a bigger-than-expected opportunity for module vendors.

Here’s the quote:

“According to Mobile Experts, connectivity module revenue will grow from about $1B in 2016 to about $33B in 2021, with volume growth in multiple markets clearly overcoming the drop in device pricing. At a semiconductor level, the revenue for MCUs and RF solutions will grow to more than $1B by 2021.”

33x better performance by module vendors vs. semiconductor vendors means … module vendors have more upside than I thought. For what it’s worth, the modules we work with don’t have a 33x price delta (in 2017) over the sum of individual components, etc. but the 33x could be driven by:

  • A) Inflated cellular module pricing,
  • B) expected cost reductions in components vs. module retail pricing, or
  • C) higher than expected heterogeneity of use cases and customer volumes that drive thousands of small deployments where (overpriced) modules are the preferred go-to-market route vs. individual components.

I’d go with option C) if I had to pick …

A side note: this report claims to use a bottoms-up forecasting methodology for the LPWAN market based, it appears, on face-to-face interviews and other exploratory methods. After the cascade of IoT forecasts showing billions or trillions of devices by next year — or maybe it was 2020— it’s refreshing to see this approach.