Impinj: Can A Chip Company Build A Competitive Moat?

| About: Impinj, Inc. (PI)
This article is now exclusive for PRO subscribers.


Impinj is the leader in selling RAIN-based RFID chips.

It has recently attracted attention due to rumors regarding its involvement with Amazon and with the Amazon Go automated grocery store.

Its product offering transcends chips and includes readers and gateways and an operating system to facilitate a complete end-user solution in many different user categories.

The company is still tiny with sales this year just crossing $100 million and a small but growing level of EPS.

Its valuation is built on achieving growth of well above the levels that make up the consensus, its relation with Amazon and its potential IoT play.

What is Impinj?

No, that is not a question to stump Jeopardy viewers, although it would have stumped me until just a few days ago. The word is supposed to be based on the representation of "Impact Ionized hot electron injection." Still wondering what it means? Well, you are hardly alone. Wondering how Impinj (NASDAQ:PI) stands for ionizing something - I won't argue the point. Those genius types, and I suppose the founders of this company must fall in that classification, never seem to realize the consequences of their activities when it comes to communicating with us mere mortals.

At one level, Impinj is a company that makes a specific class of RFID solutions that include chips, readers, antennas, accessories and software. The company's chip technology is most frequently referred to as RAIN, which is a global alliance promoting the universal adoption of UHF RFID technology in a way similar to other wireless technology organizations. It is yet another acronym standing for RAdio frequency IdentificatioN - and it is intended to represent the link between UHF RFID and the cloud where the data collected can be stored, managed and shared.

Impinj has been a hot IPO that was up by 143% from its offer price when I started writing this article. Of course, as can sometimes happen, the best laid plans of mice and men didn't quite work out. Impinj shares are very hot these days and I have been under the weather - something of a chill I suppose. The shares were up 11% yesterday on renewed Amazon (NASDAQ:AMZN) speculation although they have backed off a bit this morning.

In writing for SA, I simply do not try to precisely time an entry point into a name. I wish the shares hadn't spiked yesterday. But if the concept works, there is plenty of runway to come as I will explain later on in this article. Some readers, if they are intrigued by the name, might find it prudent to see if there is some kind of a pullback in the share price. But, of course, there are likely to be continuing rumors regarding a potential partnership with Amazon as that company gets ready to potentially roll out its Amazon Go initiative to create a fully automated grocery experience for consumers. I do not think anyone really knows how the relationship between Impinj and Amazon might develop and what the size and profitability of the relationship might be for Impinj. I do think that any kind of a press release regarding a partnership would have a material impact on Impinj's valuation, and at this point, the valuation of Impinj shares is predicated, in part, on a favorable business evolution with Amazon. Amazon is a big company and it goes down lots of technology paths simultaneously and I certainly have no specific information that Impinj is one of Amazon's chosen partners for Amazon Go or for anything else. Certainly, Impinj is a logical partner for Amazon's technology. The companies are said to have had an historic relationship, they are both located in Seattle and Amazon sells Impinj readers. Of course, Amazon sells just about everything else. Amazon recently joined the RAIN alliance of which Impinj is a leading component.

Really, the reason I think it is more than a bit likely that the two companies are working together is several recent Impinj product announcements. In particular, Impinj has been developing a technology called "always-on" which allows retailers to track inventory data in real time. That is one of the more significant components of the Amazon Go technology. In an Amazon Go deployment, using the new software that Impinj has developed, management could keep track of its inventory in real time and know where items are physically located. Developing the specifics needed specifically for Amazon's pilot is said to be easier with this technology than with predecessor solutions.

While I do think it is more likely than not that Impinj and Amazon are working together and probably working together on Amazon Go, that doesn't mean that Amazon Go will be widely rolled out - although I imagine it will. And it doesn't mean that Impinj will wind up making all of the hardware for Amazon Go or that its sales of that hardware will be particularly profitable. On the other hand, Impinj technology has use cases for Amazon far beyond the Amazon Go stores. When RFID first emerged, it was thought that it would find its most significant market in warehouse automation. A dozen years on, that still seems likely and Amazon has loads of fulfillment centers that it is constantly tweaking to bring down costs.

Amazon Go and its technologies will almost certainly have teething problems, I imagine. As I comment below, Disney (NYSE:DIS) is working on a similar technology but its accuracy is not adequate for a high-volume application like grocery stores. And then there is the sad story of Nuance (NASDAQ:NUAN) and Siri. A few years ago, Apple rolled out Siri, the voice component of its iPhone. Investors speculated that Siri was based on Nuance technology and so it proved to be. Unfortunately for Nuance, all it wound up getting was a large, multi-year implementation contract that these days is less than 2% of revenues (I think). The technology industry is full of such events and it often can happen that much transpires between getting served the meal and the enjoyment of the dinner. I assume that the people who run Impinj are all too familiar with that sorry scenario, but it is something that investors would do well to at least consider. Some observers believe that Amazon could become a customer that adds 3-5% to the growth rate of Impinj or more if the technology is deployed in Amazon fulfillment centers.

A dozen years ago, the introduction of RFID at reasonable prices was going to change the world and this company was supposedly working with Wal-Mart (NYSE:WMT) and other high volume retailers. RFID was mercilessly hyped and yet it never really revolutionized the retailing world or dramatically enhancing retailing productivity and Impinj has revenues forecast next year to be all of $140 million. I think it is necessary for investors to understand that the adoption of a new technology can have many twists and turns and it doesn't always benefit the pioneers in a straightforward fashion…if at all.

That being said, Impinj has invented a new and seemingly a better mousetrap and there are certainly straws suggesting a widespread adoption of the technology. Certainly, what Impinj is selling can be readily characterized as disruptive and the technology has the opportunity to disrupt numerous industries and to drive productivity at a meaningful rate. I think an article outlining the background of what Impinj does and presenting the case for the company as an investment is worthwhile for readers looking for that next big thing. No guarantees, but lots of pregnant possibilities.

How Impinj got here

As mentioned above, Impinj is a company with a significant history, and a background of technology that is just now starting to show some level of financial success. Its founder and CEO is Chris Diorio. Dr. Diorio studied under Carver Mead at Caltech where they developed some of the initial technology that is used in their current set of solutions. Dr. Mead is one of the luminaries in the semiconductor world and is still engaged with Impinj at some level even at age 82.

The company apparently flirted with greatness on a few occasions over the course of its history and Bear Steans reported that Wal-Mart had issued contracts to Impinj and to Allen Technology to supply some of the hardware for a total of 15,000 RFID readers that were needed for Wal-Mart distribution centers at that time. The report was never confirmed.

As mentioned in the company's prospectus, the technology that this company sells, RAIN, has had an uneven adoption history with spurts of growth followed by slowdowns in the adoption of the solution. In the prospectus, the company speculates that the cause for the latest cycle had to do with a patent dispute that is now settled. I think that a more likely explanation is that this was a technology that never got its costs to a level that sparked widespread adoption. The tags still cost several cents per item which meant that item-level tracking for everyday products has remained too expensive. Retail apparel, which costs far more than typical everyday items, has been the largest use case for the technology with customers such as Macy's (NYSE:M) broadly adopting RAIN standards for the apparel items that it sells.

The company tried to go public as far back as 2011 but the reception to its proposed offering was not positive given the uncertain adoption of RAIN at that point and the dispute with Round Rock regarding its IP. The following year the company raised $21 million which carried it through to the recent IPO. Overall, the company raised $132 million in private equity from VCs such as Intel Capital, AllianceBernstein and Madrona.

The success that the shares have enjoyed subsequent to the very modestly sized IPO allowed the company to launch a secondary offering which was completed at the end of November. The company wound up selling a bit over 4 million shares at a price of $27 share. The prospectus suggested that some of the proceeds of the offering might be used for tuck-in acquisitions and that seems a likely strategy at this point. Of the shares that were sold, the company offered 1.5 million, VCs offered 2.5 million and the CEO offered 800,000 out of his total ownership of 1.17 million. There is a significant lock-up expiration that is effective on January 11th.

The current hope/expectation for this technology relates to the development of the Internet of Things (IoT). In order to enable the Iot there is a need to identify objects and link them to the Internet and the use of passive ultrahigh frequency tags are emerging as a likely standard. The RFID technology that Impinj offers relates to obtaining information as to what is happening to a specific object located at a particular point at an exact time. When I said that I wouldn't be surprised see Amazon and Impinj partnering for the Amazon Go store technology, part of my speculation relates to the fact that in an Amazon Go store it is crucial to be able to track the kind of data that the Impinj platform can generate.

After the hype cycle that probably peaked in the middle years of the last decade, the technology hit a trough between 2008 and 2011 according to Raghu Das, CEO of IDTechEx, a market research firm whose research is quoted in the Impinj prospectus. According to the research compiled by Das, the growth in tags sold was a bit greater than 30%. RAIN tags were about 76% of the total volume of tags shipped and IDC is forecasting that the volume of RAIN tags sold is going to reach 20 billion by the end of the decade which a growth rate of 27%.

Frost & Sullivan suggests that the TAM for RFID within the retail industry will be $5.3 billion by 2020, a CAGR of 39%. Another market research vendor called Transparency forecasts that the TAM for RFID within the healthcare space will be of comparable size, growing at a CAGR of 14%.

There is a huge amount of runway available for a company the size of Impinj selling solutions in markets with TAMs such as those indicated. The question, of course, is how realistic those TAMs are going to be and if this time the RAIN will really fall from the clouds. How terrible a pun is that?

When investors are looking for heavy RAIN - or the case for Impinj

In looking at the probabilities of success for any technology, there always need to be a confluence of events. Self-evidently, the internet could never have developed without significant advances in compute power and communications technology. Automobiles needed dramatic advances in internal combustion engines coupled with breakthrough production techniques before they could become ubiquitous. Autonomous cars are developing now because the costs of the lasers and other devices necessary to avoid collisions have dropped to a point where redundant systems are economically feasible. Much time and effort have gone into the analysis of why railways developed when they did. But it took the confluence of light and powerful steam engines, the development of metallurgy so that reasonably priced rails could be produced and the development of civil engineering so that bridges could be built over wide rivers to make the development of dense railway networks possible. People understood the principles of vehicles traveling on rails long before railways became practical.

The initial hype for RFID was misplaced not because RFID doesn't do what users need done, but it needs a complete infrastructure in order to produce satisfactory results. The original RFID was ahead of its time because the costs of the chips were too high and the value they provided users when compared to a mature bar code technology was too small. At this point, RFID has matured significantly and it is being used far beyond the bounds of the retailing industry.

For example, Disney Research demonstrated an RFID technology called IDSense that allowed it to simultaneously track 20 objects in a room and infer four classes of movement with 93% accuracy. In the case of Disney, of course, the technology is being used for theme park identity bracelets, hotel room access cards and rental car key fobs.

Impinj is still very dependent on selling chips that are used in apparel classification but it is also focused on selling its solutions in the health care space, and is starting to achieve acceptance in automotive, industrial and manufacturing, consumer experience and even food. The number of use cases that are likely to be developed over the coming years seems likely to be far beyond those in favor at this writing.

Currently, one of the hottest space for RFID deployment is that of the airline industry and the use of the technology to track bags. This past year, Delta (NYSE:DAL) was the primary customer to deploy RFID bag tracking solutions but there are many airlines that seem likely to want to roll out similar technologies to minimize lost bag problems. The current state of the technology is such that the chip is always on and needs no battery to tell trackers where it is located. I'm not sure how the Delta roll-out has gone, but apparently the Delta roll-out has achieved a level of success that the other airlines are engaging with Impinj in terms of considering the technology. How large the airline opportunity actually might be? What the cadence of a roll-out might be? And whether Impinj will face serious competitors as the technology matures are items that really can't be determined precisely at this point. I do not think it is possible to determine a long-term CAGR for this company other than to speculate it is likely to be high - and greater than the mid-20s range that it talked about during the course of its IPO.

Competition - can Impinj maintain its competitive moat?

There has never been a tech IPO without that kind of question. Impinj makes chips. I am sure that it will be said that chips are a known commodity. Eventually, Impinj will lose market share and its margins will implode. The Chinese will make an equivalent version of RAIN chips and sell them for far less than the prices of the Impinj chip. These are all arguments that are likely to be heard and more so if the shares continue to perform well.

I think it fair to suggest that the old bromide of a rising tide lifting all boats is an apposite rejoinder to much of the above. If RFID/RAIN technology is now about to fulfill its promise, there will be plenty of demand and competition is not likely to be such that it destroys the ability of Impinj to meet its targets with regards to growth. As cited earlier in this article (In the Impinj S-1), the expectations for end-point growth for RAIN are in the range of 27-29% over the next several years. That is probably a good starting point from which to model demand. But the forecasts do not really include the potential from sales of this company's complete platform and that is a very nascent business at this point.

I think it should be obvious that the secret sauce that Impinj sells is not its chip. The company sells a platform that includes its chip to be sure, but also includes offerings that are based on its readers and gateways. These are high value products and offer far higher barriers to competition. The company introduced its software in 2015. It is the software that presumably will differentiate the capabilities that Impinj is able to offer its users for the foreseeable future.

Chips do tend to be commodities and often carry commodity margins. Complex devices such as gateways are not commodities and software, by its nature, is hard to replicate.

Because of its leadership in the space, Impinj is developing an ecosystem. That is an over-hyped concept if ever there was one. But just because it is over-hyped doesn't mean it isn't real. The company has multiple joint solutions with various system integrators that bring the value of its technology to end-users in very specific ways. Many of these solutions will never produce significant volumes, some of them will be singles and a few of them will be home runs. But the point is that the Impinj solution and its operating system are the backbone of the solution and are not subject to the commodity trends that impact most chip manufacturers.

The entire essence of this company is to sell solutions and not chips and that is its differentiation and its moat and seems likely to continue. Obviously, it has to sell billions of end-points in order to drive demand for the balance of its technology, but part of user decisions regarding end-points is the balance of the solution.

The company wrote in its S-1 filing that it has a greater than 60% market share currently. It is a useful metric to track and market dominance does lead to above average margins. I am sure that the growth of the market is going to entice additional competitors over time. But in some ways, the kind of competition I foresee relates to a race between the adoption of the Impinj platform and the availability of commodity chips. If the relationship with Amazon blossoms, it will be a signpost, if nothing more, that Impinj is winning that race.


The question of how to value a company like this is intriguing. The company is only followed by five analysts all of whom were part of the underwriting group and all of whom decided the shares were a buy. The share price has now sailed through the consensus valuation as the speculation regarding Amazon grows more insistent.

The company reported non-GAAP EPS of $.11 last quarter. Revenue growth accelerated significantly to 52%, up from 36% growth the prior quarter and 9% above the consensus. Stock-based comp was negligible for the quarter and accounted for $.01 of reported EPS. The company provided a guidance range of $.04-.11 for current quarter EPS with revenue growth of around 43%.

Year to date, the company experienced a significant cash burn of $9.5 million. The burn was entirely a function of the increase in receivables and inventories. The increase in these categories was a function of the extremely rapid growth of the company's revenues which substantially exceeded expectations last quarter. In the last quarter, the growth in inventories slowed down while the growth in receivables increased. That is what one ought to expect from a company in this stage of growth.

Analyst estimates are congruent with guidance. Currently, expectations for 2017 are for growth of 25% in revenues to $140 million and for EPS of $.27. No one is buying Impinj shares currently because they think that revenues are going to growing at 25% or because they expect the company to earn $.27.

The company has about 20 million shares outstanding subsequent to the recent secondary. Subsequent to the secondary, it has about $80 million of net cash. So, its enterprise value, after yesterday's run-up is about $700 million.

Will Amazon add to the company's growth and how much? Will the company's success with Amazon have some kind of drag-along impact and open the doors to additional use cases in areas that are not currently envisioned? Will the success of Delta with its RFID deployment contribute to rapid growth in the airline vertical? I would be naive and arrogant if I claimed to know the answers to these questions and certainly not at any level of detail necessary to forecast the growth and profitability of this company.

If Impinj were a software company, its valuation would be low. If Impinj is to be valued as a semiconductor vendor, its valuation is very high. It is surely a hybrid and a hybrid with some unique attributes. I have personally not acquired shares in the company, although I am considering adding it to my portfolio. If I miss it there is an Amazon announcement, I will just have to live with that. The word Unicorn is one whose meaning has evolved massively over the years from that of being a mythical creature with a horn and as the emblem of Scotland to being a private company with a billion dollars of valuation. Fortune listed no fewer than 174 such companies earlier this year. Impinj is not a unicorn by either definition. I used to think of a unicorn as something unique in a positive fashion and that is how I would describe Impinj. It is a very unique combination of semiconductor technology and software that can produce dramatic productivity gains for users. I think the company is likely to produce positive alpha when looked back from a perspective of a year or two. I offer no perspective as to how to trade the name.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in PI over the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.